In this article, I will present to you the most sophisticated optimization algorithms in Deep Learning that allow neural networks to learn faster and achieve better performance. These algorithms are Stochastic Gradient Descent with Momentum...

In this article, I will present to you the most sophisticated optimization algorithms in Deep Learning that allow neural networks to learn faster and achieve better performance.

These algorithms are Stochastic Gradient Descent with Momentum, AdaGrad, RMSProp, and Adam Optimizer.

Table of Content- Why do we need better optimization Algorithms?
- Stochastic Gradient Descent with Momentum
- AdaGrad
- RMSProp
- Adam Optimizer
- What is the best Optimization Algorithm for Deep Learning?

To train a neural network model, we must define a loss function in order to measure the difference between our model predictions and the label that we want to predict. What we are looking for is a certain set of weights, with which the neural network can make an accurate prediction, which automatically leads to a lower value of the loss function.

I think you must know by now, that the mathematical method behind it is called gradient descent.

Eq. 1 Gradient Descent for parameters θ with loss function L.

In this technique (Eq.1), we must calculate the gradient of the loss function ** L** with respect to the weights (or parameters

By periodically applying the gradient descent to the weights, we will eventually arrive at the optimal weights that minimize the loss function and allow the neural network to make better predictions.

So far the theory.

Do not get me wrong, gradient descent is still a powerful technique. In practice, however, this technique may encounter certain problems during training that can slow down the learning process or, in the worst case, even prevent the algorithm from finding the optimal weights

These problems were on the one hand **saddle points** and **local minima** of the loss function, where the loss function becomes flat and the gradient goes to zero:

Fig. 1 Saddle Points and Local Minima

A gradient near zero does not improve the weight parameters and prevents the entire learning process.

On the other hand, even if we have gradients that are not close to zero, the values of these gradients calculated for different data samples from the training set may vary in value and direction. We say that the gradients are noisy or have a lot of variances. This leads to a zigzag movement towards the optimal weights and can make learning much slower:

Fig. 3 Example of zig-zag movements of noisy gradients.

In the following article, we are going to learn about more sophisticated gradient descent algorithms. All of these algorithms are based on the regular gradient descent optimization that we have come to know so far. But we can extend this regular approach for the weight optimization by some mathematical tricks to build even more effective optimization algorithms that allow our neural network to adequately handle these problems, thereby learning faster and to achieve a better performance

2. Stochastic Gradient Descent with MomentumThe first of the sophisticated algorithms I want to present you is called stochastic gradient descent with momentum.

Eq. 2 Equations for stochastic gradient descent with momentum.

On the left side in Eq. 2, you can see the equation for the weight updates according to the regular stochastic stochastic gradient descent. The equation on the right shows the rule for the weight updates according to the SGD with momentum. The momentum appears as an additional term ** ρ** times

Intuitively speaking, by adding this momentum term we let our gradient to build up a kind of velocity ** v** during training. The velocity is the running sum of gradients weighted by

** ρ** can be considered as friction that slows down the velocity a little bit. In general, you can see that the velocity builds up over time. By using the momentum term saddle points and local minima become less dangerous for the gradient. Because step sizes towards the global minimum now don’t depend only on the gradient of the loss function at the current point, but also on the velocity that has built up over time.

In other words, we are moving more towards the direction of velocity than towards the gradient at a certain point.

If you want to have a physical representation of the stochastic gradient descent with momentum think about a ball that rolls down a hill and builds up velocity over time. If this ball reaches some obstacles on its way, such as a hole or a flat ground with no downward slope, the velocity ** v** would give the ball enough power to roll over these obstacles. In this case, the flat ground and the hole represent saddle points or local minima of a loss function.

In the following video (Fig. 4), I want to show you a direct comparison of regular stochastic gradient descent and stochastic gradient descent with momentum term. Both algorithms are trying to reach the global minimum of the loss function which lives in a 3D space. Please note how the momentum term makes the gradients to have less variance and fewer zig-zags movements.

Fig. 4 SGD vs. SGD with Momentum

In general, the momentum term makes converges towards optimal weights more stable and faster.

3. AdaGradAnother optimization strategy is called AdaGrad. The idea is that you keep the running sum of squared gradients during optimization. In this case, we have no momentum term, but an expression ** g** that is the sum of the squared gradients.

Eq. 3 Parameter update rule for AdaGrad.

When we update a weight parameter, we divide the current gradient by the root of that term ** g**. To explain the intuition behind AdaGrad, imagine a loss function in a two-dimensional space in which the gradient of the loss function in one direction is very small and very high in the other direction.

Summing up the gradients along the axis where the gradients are small causes the squared sum of these gradients to become even smaller. If during the update step, we divide the current gradient by a very small sum of squared gradients ** g**, the result of that division becomes very high and vice versa for the other axis with high gradient values.

As a result, we force the algorithm to make updates in any direction with the same proportions.

This means that we accelerate the update process along the axis with small gradients by increasing the gradient along that axis. On the other hand, the updates along the axis with the large gradient slow down a bit.

However, there is a problem with this optimization algorithm. Imagine what would happen to the sum of the squared gradients when training takes a long time. Over time, this term would get bigger. If the current gradient is divided by this large number, the update step for the weights becomes very small. It is as if we were using very low learning that becomes even lower the longer the training goes. In the worst case, we would get stuck with AdaGrad and the training would go on forever.

4. RMSPropThere is a slight variation of AdaGrad called RMSProp that addresses the problem that AdaGrad has. With RMSProp we still keep the running sum of squared gradients but instead of letting that sum grow continuously over the period of training we let that sum actually decay.

Eq. 4 Update rule for RMS Prop.

In RMSProp we multiply the sum of squared gradients by a decay rate **α** and add the current gradient weighted by (1- **α)**. The update step in the case of RMSProp looks exactly the same as in AdaGrad where we divide the current gradient by the sum of squared gradients to have this nice property of accelerating the movement along the one dimension and slowing down the movement along the other dimension.

Let’s see how RMSProp is doing in comparison with SGD and SGD with momentum in finding the optimal weights.

Fig. 5 SGD vs. SGD with Momentum vs. RMS Prop

Although SGD with momentum is able to find the global minimum faster, this algorithm takes a much longer path, that could be dangerous. Because a longer path means more possible saddle points and local minima. RMSProp, on the other hand, goes straight towards the global minimum of the loss function without taking a detour.

5. Adam OptimizerSo far we have used the moment term to build up the velocity of the gradient to update the weight parameter towards the direction of that velocity. In the case of AdaGrad and RMSProp, we used the sum of the squared gradients to scale the current gradient, so we could do weight updates with the same ratio in each dimension.

These two methods seemed pretty good ideas. Why do not we just take the best of both worlds and combine these ideas into a single algorithm?

This is the exact concept behind the final optimization algorithm called Adam, which I would like to introduce to you.

The main part of the algorithm consists of the following three equations. These equations may seem overwhelming at first, but if you look closely, you’ll see some familiarity with previous optimization algorithms.

Eq. 5 Parameter update rule for Adam Optimizer

The first equation looks a bit like the SGD with momentum. In the case, the term would be the velocity and the friction term. In the case of Adam, we call the first momentum and is just a hyperparameter.

The difference to SGD with momentum, however, is the factor (1- ** β1**), which is multiplied with the current gradient.

The second part of the equations, on the other hand, can be regarded as RMSProp, in which we are keeping the running sum of squared gradients. Also, in this case, there is the factor (1-** β2**) which is multiplied with the squared gradient.

The term in the equation is called the second momentum and is also just a hyperparameter. The final update equation can be seen as a combination of RMSProp and SGD with Momentum.

So far, Adam has integrated the nice features of the two previous optimization algorithms, but here’s a little problem, and that’s the question of what happens in the beginning.

At the very first time step, the first and second momentum terms are initialized to zero. After the first update of the second momentum, this term is still very close to zero. When we update the weight parameters in the last equation, we divide by a very small second momentum term ** v**, resulting in a very large first update step.

This first very large update step is not the result of the geometry of the problem, but it is an artifact of the fact that we have initialized the first and second momentum to zero. To solve the problems of large first update steps, Adam includes a correction clause:

Eq. 6 Bias Correction for Adam Optimizer

You can see that after the first update of the first and second momentum and we make an unbiased estimate of these momentums by taking into account the current time step. These correction terms make the values of the first and second momentum to be higher in the beginning than in the case without the bias correction.

As a result, the first update step of the neural network parameters does not get that large and we don’t mess up our training in the beginning. The additional bias corrections give us the full form of Adam Optimizer.

Now, let us compare all algorithms with each other in terms of finding the global minimum of the loss function:

Fig. 6 Comparison of all optimization algorithms.

Finally, we can discuss the question of what the best gradient descent algorithm is.

In general, a normal gradient descent algorithm is more than adequate for simpler tasks. If you are not satisfied with the accuracy of your model you can try out RMSprop or add a momentum term to your gradient descent algorithms.

But in my experience the best optimization algorithm for neural networks out there is Adam. This optimization algorithm works very well for almost any deep learning problem you will ever encounter. Especially if you set the hyperparameters to the following values:

*β1=0.9**β2=0.999*- Learning rate = 0.001–0.0001

… this would be a very good starting point for any problem and virtually every type of neural network architecture I’ve ever worked with.

That’s why Adam Optimizer is my default optimization algorithm for every problem I want to solve. Only in very few cases do I switch to other optimization algorithms that I introduced earlier.

In this sense, I recommend that you always start with the Adam Optimizer, regardless of the architecture of the neural network of the problem domain you are dealing with.

Cheat Sheets for AI, Neural Networks, Machine Learning, Deep Learning & Big Data

Downloadable PDF of Best AI Cheat Sheets in Super High DefinitionLet’s begin.

Cheat Sheets for AI, Neural Networks, Machine Learning, Deep Learning & Data Science in HD

Neural Networks Cheat Sheets

Neural Networks Basics Cheat Sheet

An Artificial Neuron Network (ANN), popularly known as Neural Network is a computational model based on the structure and functions of biological neural networks. It is like an artificial human nervous system for receiving, processing, and transmitting information in terms of Computer Science.

- Input Layer (All the inputs are fed in the model through this layer)
- Hidden Layers (There can be more than one hidden layers which are used for processing the inputs received from the input layers)
- Output Layer (The data after processing is made available at the output layer)

Neural Networks Graphs Cheat Sheet

Graph data can be used with a lot of learning tasks contain a lot rich relation data among elements. For example, modeling physics system, predicting protein interface, and classifying diseases require that a model learns from graph inputs. Graph reasoning models can also be used for learning from non-structural data like texts and images and reasoning on extracted structures.

Machine Learning Cheat Sheets

Machine Learning with Emojis Cheat Sheet

Scikit Learn Cheat Sheet

Scikit-learn is a free software machine learning library for the Python programming language. It features various classification, regression and clustering algorithms including support vector machines is a simple and efficient tools for data mining and data analysis. It’s built on NumPy, SciPy, and matplotlib an open source, commercially usable — BSD license

Scikit-learn Algorithm Cheat Sheet

This machine learning cheat sheet will help you find the right estimator for the job which is the most difficult part. The flowchart will help you check the documentation and rough guide of each estimator that will help you to know more about the problems and how to solve it.

If you like these cheat sheets, you can let me know here.### Machine Learning: Scikit-Learn Algorythm for Azure Machine Learning Studios

Scikit-Learn Algorithm for Azure Machine Learning Studios Cheat Sheet

Data Science with Python Cheat Sheets

TensorFlow Cheat Sheet

TensorFlow is a free and open-source software library for dataflow and differentiable programming across a range of tasks. It is a symbolic math library, and is also used for machine learning applications such as neural networks.

If you like these cheat sheets, you can let me know here.### Data Science: Python Basics Cheat Sheet

Python Basics Cheat Sheet

Python is one of the most popular data science tool due to its low and gradual learning curve and the fact that it is a fully fledged programming language.

PySpark RDD Basics Cheat Sheet

“At a high level, every Spark application consists of a *driver program* that runs the user’s `main`

function and executes various *parallel operations* on a cluster. The main abstraction Spark provides is a *resilient distributed dataset* (RDD), which is a collection of elements partitioned across the nodes of the cluster that can be operated on in parallel. RDDs are created by starting with a file in the Hadoop file system (or any other Hadoop-supported file system), or an existing Scala collection in the driver program, and transforming it. Users may also ask Spark to *persist* an RDD in memory, allowing it to be reused efficiently across parallel operations. Finally, RDDs automatically recover from node failures.” via Spark.Aparche.Org

NumPy Basics Cheat Sheet

NumPy is a library for the Python programming language, adding support for large, multi-dimensional arrays and matrices, along with a large collection of high-level mathematical functions to operate on these arrays.

***If you like these cheat sheets, you can let me know *****here.**

Bokeh Cheat Sheet

“Bokeh is an interactive visualization library that targets modern web browsers for presentation. Its goal is to provide elegant, concise construction of versatile graphics, and to extend this capability with high-performance interactivity over very large or streaming datasets. Bokeh can help anyone who would like to quickly and easily create interactive plots, dashboards, and data applications.” from Bokeh.Pydata.com

Karas Cheat Sheet

Keras is an open-source neural-network library written in Python. It is capable of running on top of TensorFlow, Microsoft Cognitive Toolkit, Theano, or PlaidML. Designed to enable fast experimentation with deep neural networks, it focuses on being user-friendly, modular, and extensible.

Padas Basics Cheat Sheet

Pandas is a software library written for the Python programming language for data manipulation and analysis. In particular, it offers data structures and operations for manipulating numerical tables and time series. It is free software released under the three-clause BSD license.

If you like these cheat sheets, you can let me know here.### Pandas Cheat Sheet: Data Wrangling in Python

Pandas Cheat Sheet: Data Wrangling in Python

The term “data wrangler” is starting to infiltrate pop culture. In the 2017 movie Kong: Skull Island, one of the characters, played by actor Marc Evan Jackson is introduced as “Steve Woodward, our data wrangler”.

Data Wrangling with Pandas Cheat Sheet

- Although many fundamental data processing functions exist in R, they have been a bit convoluted to date and have lacked consistent coding and the ability to easily
*flow*together → leads to difficult-to-read nested functions and/or*choppy*code. - R Studio is driving a lot of new packages to collate data management tasks and better integrate them with other analysis activities → led by Hadley Wickham & the R Studio team → Garrett Grolemund, Winston Chang, Yihui Xie among others.
- As a result, a lot of data processing tasks are becoming packaged in more cohesive and consistent ways → leads to:
- More efficient code
- Easier to remember syntax
- Easier to read syntax” via Rstudios

Data Wrangling with ddyr and tidyr Cheat Sheet

If you like these cheat sheets, you can let me know here.### Data Science: Scipy Linear Algebra

Scipy Linear Algebra Cheat Sheet

SciPy builds on the NumPy array object and is part of the NumPy stack which includes tools like Matplotlib, pandas and SymPy, and an expanding set of scientific computing libraries. This NumPy stack has similar users to other applications such as MATLAB, GNU Octave, and Scilab. The NumPy stack is also sometimes referred to as the SciPy stack.[3]

Matplotlib Cheat Sheet

Matplotlib is a plotting library for the Python programming language and its numerical mathematics extension NumPy. It provides an object-oriented APIfor embedding plots into applications using general-purpose GUI toolkits like Tkinter, wxPython, Qt, or GTK+. There is also a procedural “pylab” interface based on a state machine (like OpenGL), designed to closely resemble that of MATLAB, though its use is discouraged. SciPy makes use of matplotlib.

Pyplot is a matplotlib module which provides a MATLAB-like interface matplotlib is designed to be as usable as MATLAB, with the ability to use Python, with the advantage that it is free.

Data Visualization with ggplot2 Cheat Sheet

Big-O Cheat Sheet

Special thanks to DataCamp, Asimov Institute, RStudios and the open source community for their content contributions. You can see originals here:

Big-O Algorithm Cheat Sheet: http://bigocheatsheet.com/

Bokeh Cheat Sheet: https://s3.amazonaws.com/assets.datacamp.com/blog_assets/Python_Bokeh_Cheat_Sheet.pdf

Data Science Cheat Sheet: https://www.datacamp.com/community/tutorials/python-data-science-cheat-sheet-basics

Data Wrangling Cheat Sheet: https://www.rstudio.com/wp-content/uploads/2015/02/data-wrangling-cheatsheet.pdf

Data Wrangling: https://en.wikipedia.org/wiki/Data_wrangling

Ggplot Cheat Sheet: https://www.rstudio.com/wp-content/uploads/2015/03/ggplot2-cheatsheet.pdf

Keras Cheat Sheet: https://www.datacamp.com/community/blog/keras-cheat-sheet#gs.DRKeNMs

Keras: https://en.wikipedia.org/wiki/Keras

Machine Learning Cheat Sheet: https://ai.icymi.email/new-machinelearning-cheat-sheet-by-emily-barry-abdsc/

Machine Learning Cheat Sheet: https://docs.microsoft.com/en-in/azure/machine-learning/machine-learning-algorithm-cheat-sheet

ML Cheat Sheet:: http://peekaboo-vision.blogspot.com/2013/01/machine-learning-cheat-sheet-for-scikit.html

Matplotlib Cheat Sheet: https://www.datacamp.com/community/blog/python-matplotlib-cheat-sheet#gs.uEKySpY

Matpotlib: https://en.wikipedia.org/wiki/Matplotlib

Neural Networks Cheat Sheet: http://www.asimovinstitute.org/neural-network-zoo/

Neural Networks Graph Cheat Sheet: http://www.asimovinstitute.org/blog/

Neural Networks: https://www.quora.com/Where-can-find-a-cheat-sheet-for-neural-network

Numpy Cheat Sheet: https://www.datacamp.com/community/blog/python-numpy-cheat-sheet#gs.AK5ZBgE

NumPy: https://en.wikipedia.org/wiki/NumPy

Pandas Cheat Sheet: https://www.datacamp.com/community/blog/python-pandas-cheat-sheet#gs.oundfxM

Pandas: https://en.wikipedia.org/wiki/Pandas_(software)

Pandas Cheat Sheet: https://www.datacamp.com/community/blog/pandas-cheat-sheet-python#gs.HPFoRIc

Pyspark Cheat Sheet: https://www.datacamp.com/community/blog/pyspark-cheat-sheet-python#gs.L=J1zxQ

Scikit Cheat Sheet: https://www.datacamp.com/community/blog/scikit-learn-cheat-sheet

Scikit-learn: https://en.wikipedia.org/wiki/Scikit-learn

Scikit-learn Cheat Sheet: http://peekaboo-vision.blogspot.com/2013/01/machine-learning-cheat-sheet-for-scikit.html

Scipy Cheat Sheet: https://www.datacamp.com/community/blog/python-scipy-cheat-sheet#gs.JDSg3OI

SciPy: https://en.wikipedia.org/wiki/SciPy

TesorFlow Cheat Sheet: https://www.altoros.com/tensorflow-cheat-sheet.html

Data Science, Machine Learning, Deep Learning, and Artificial intelligence are really hot at this moment and offering a lucrative career to programmers with high pay and exciting work.

Data Science, Machine Learning, Deep Learning, and Artificial intelligence are really hot at this moment and offering a lucrative career to programmers with high pay and exciting work.

It's a great opportunity for programmers who are willing to learn these new skills and upgrade themselves and want to solve some of the most interesting real-world problems.

It's also important from the job perspective because Robots and Bots are getting smarter day by day, thanks to these technologies and most likely will take over some of the jobs which many programmers do today.

Hence, it's important for software engineers and developers to upgrade themselves with these skills. Programmers with these skills are also commanding significantly higher salaries as data science is revolutionizing the world around us.

You might already know that the Machine learning specialist is one of the top paid technical jobs in the world. However, most developers and IT professionals are yet to learn this valuable set of skills.

For those, who don't know what is a Data Science, Machine learning, or deep learning, they are very related terms with all pointing towards machine doing jobs which is only possible for humans till date and analyzing the huge set of data collected by modern day application.

Data Science, in particular, is a combination of concepts such as machine learning, visualization, data mining, programming, data mugging, etc.

If you have some programming experience then you can learn Python or Rto make your carer as a Data Scientist.

There are a lot of popular scientific Python libraries such as Numpy, Scipy, Scikit-learn, Pandas, which is used by Data Scientist for analyzing data.

To be honest with you, I am also quite new to Data Science and Machine learning world but I have been spending some time from last year to understand this field and have done some research in terms of best resources to learn machine learning, data science, etc.

I am sharing all those resources in a series of a blog post like this. Earlier, I have shared some courses to learn **TensorFlow**, one of the most popular machine-learning library and today I'll share some more to learn these technologies.

These are a combination of both free and paid resource which will help you to understand key data science concepts and become a Data Scientist. Btw, I'll get paid if you happen to buy a course which is not free.

Here is my list of some of the best courses to learn Data Science, Machine learning, and deep learning using Python and R programming language. As I have said, Data Science and machine learning work very closely together, hence some of these courses also cover machine learning.

If you are still on fence with respect to choosing Python or R for machine learning, let me tell you that both Python and R are a great language for Data Analysis and have good APIs and library, hence I have included courses in both Python and R, you can choose the one you like.

I personally like Python because of its versatile usage, it's the next best in my list of language after Java. I am already using it for writing scripts and other web stuff, so it was an easy choice for me. It has also got some excellent libraries like Sci-kit Learn and TensorFlow.

Data Science is also a combination of many skills e.g. visualization, data cleaning, data mining, etc and these courses provide a good overview of all these concepts and also presents a lot of useful tools which can help you in the real world.

**Machine Learning by Andrew Ng**

This is probably the most popular course to learn machine learning provided by Stanford University and Coursera, which also provides certification. You'll be tested on each and every topic that you learn in this course, and based on the completion and the final score that you get, you'll also be awarded the certificate.

This course is free but you need to pay for certificates, if you want. Though, it does provide value to you as a developer and gives you a good understanding of the mathematics behind all the machine learning algorithms that you come up with.

I personally really like this one. Andrew Ng takes you through the course using Octave, which is a good tool to test your algorithm before making it go live on your project.

1.**Machine Learning A-Z: Hands-On Python and R --- In Data Science**

This is probably the best hands on course on Data Science and machine learning online. In this course, you will learn to create Machine Learning Algorithms in Python and R from two Data Science experts.

This is a great course for students and programmers who want to make a career in Data Science and also Data Analysts who want to level up in machine learning.

It's also good for any intermediate level programmers who know the basics of machine learning, including the classical algorithms like linear regression or logistic regression, but who want to learn more about it and explore all the different fields of Machine Learning.

**2.** **Data Science with R by Pluralsight**

Data science is the practice of transforming data into knowledge, and R is one of the most popular programming language used by data scientists.

In this course, you'll learn first learn about the practice of data science, the R programming language, and how they can be used to transform data into actionable insight.

Next, you'll learn how to transform and clean your data, create and interpret descriptive statistics, data visualizations, and statistical models.

Finally, you'll learn how to handle Big Data, make predictions using machine learning algorithms, and deploy R to production.

Btw, you would need a Pluralsight membership to get access this course, but if you don't have one you can still check out this course by taking their **10-day free Pass**, which provides 200 minutes of access to all of their courses for free.

**3.**** ****Harvard Data Science Course**

The course is a combination of various data science concepts such as machine learning, visualization, data mining, programming, data mugging, etc.

You will be using popular scientific Python libraries such as Numpy, Scipy, Scikit-learn, Pandas throughout the course.

I suggest you complete the machine learning course on course before taking this course, as machine learning concepts such as PCA (dimensionality reduction), k-means and logistic regression are not covered in depth.

But remember, you have to invest a lot of time to complete this course, especially the homework exercises are very challenging

In short, if you are looking for an online course in data science(using Python), there is no better course than Harvard's CS 109. You need some background in programming and knowledge of statistics to complete this course.

**4.** **Want to be a Data Scientist? (FREE)**

This is a great introductory course on what Data Scientist do and how you can become a data science professional. It's also free and you can get it on Udemy.

If you have just heard about Data Science and excited about it but doesn't know what it really means then this is the course you should attend first.

It's a small course but packed with big punches. You will understand what Data Science is? Appreciate the work Data Scientists do on a daily basis and differentiate the various roles in Data Science and the skills needed to perform them.

You will also learn about the challenges Data Scientists face. In short, this course will give you all the knowledge to make a decision on whether Data Science is the right path for you or not.

**5.** **Intro to Data Science by Udacity**

This is another good Introductory course on Data science which is available for free on Udacity, another popular online course website.

In this course, you will learn about essential Data science concepts e.g. Data Manipulation, Data Analysis with Statistics and Machine Learning, Data Communication with Information Visualization, and Data at Scale while working with Big Data.

This is a free course and it's also the first step towards a new career with the Data Analyst Nanodegree Program offered by Udacity.

**6.** **Data Science Certification Training --- R Programming**

The is another good course to learn Data Science with R. In this course, you will not only learn R programming language but also get some hands-on experience with statistical modeling techniques.

The course has real-world examples of how analytics have been used to significantly improve a business or industry.

If you are interested in learning some practical analytic methods that don't require a ton of maths background to understand, this is the course for you.

**7.** **Intro To Data Science Course by Coursera**

This course provides a broad introduction to various concepts of data science. The first programming exercise "Twitter Sentiment Analysis in Python" is both fun and challenging, where you analyze tons of twitter message to find out the sentiments e.g. negative, positive etc.

The course assumes that you know statistics, Python, and SQL.

Btw, It's not so good for beginners, especially if you don't know Python and SQL but if you do and have a basic understanding of Data Science then this is a great course.

8. **Python for Data Science and Machine Learning Bootcamp**

There is no doubt that Python is probably the best language, apart from R for Data Analysis and that's why it's hugely popular among Data Scientists.

This course will teach you how to use all important Python scientific and machine learning libraries Tensorflow, NumPy, Pandas, Seaborn, Matplotlib, Plotly, Scikit-Learn, Machine Learning, and many more libraries which I have explained earlier in my list of useful machine learning libraries.

It's a very comprehensive course and you will how to use the power of Python to analyze data, create beautiful visualizations, and use powerful machine learning algorithms!

**9.** **Data Science A-Z: Real-Life Data Science Exercises Included**

This is another great hands-on course on Data Science from Udemy. It promises to teach you Data Science step by step through real Analytics examples. Data Mining, Modeling, Tableau Visualization and more.

This course will give you so many practical exercises that the real world will seem like a piece of cake when you complete this course.

The homework exercises are also very thought-provoking and challenging. In short, If you love doing stuff then this is a course for you.

10. **Data Science, Deep Learning and Machine Learning with Python**

If you've got some programming or scripting experience, this course will teach you the techniques used by real data scientists and machine learning practitioners in the tech industry --- and help you to become a data scientist.

The topics in this course come from an analysis of real requirements in data scientist job listings from the biggest tech employers, that makes it even more special and useful.

That's all about some of the **popular courses to learn Data Science**. As I said, there is a lot of demand for good Data Analytics and there are not many developers out there to fulfill that demand.

It's a great chance for the programmer, especially those who have good knowledge of maths and statistics to make a career in machine learning and Data analytics. You will be awarded exciting work and incredible pay.

Other useful **Data Science and Machine Learning** resources

Top 8 Python Machine Learning Libraries

5 Free courses to learn R Programming for Machine learning

5 Free courses to learn Python in 2018

Top 5 Data Science and Machine Learning courses

Top 5 TensorFlow and Machine Learning Courses

10 Technologies Programmers Can Learn in 2018

Top 5 Courses to Learn Python Better

How a Japanese cucumber farmer is using deep learning and TensorFlow

Thanks, You made it to the end of the article ... Good luck with your Data Science and Machine Learning journey! It's certainly not going to be easy, but by following these courses, you are one step closer to becoming the Machine Learning Specialists you always wanted to be.

A step-by-step guide to setting up Python for Deep Learning and Data Science for a complete beginner

A step-by-step guide to setting up Python for Deep Learning and Data Science for a complete beginner

You can code your own Data Science or Deep Learning project in just a couple of lines of code these days. This is not an exaggeration; many programmers out there have done the hard work of writing tons of code for us to use, so that all we need to do is plug-and-play rather than write code from scratch.

You may have seen some of this code on Data Science / Deep Learning blog posts. Perhaps you might have thought: “Well, if it’s really that easy, then why don’t I try it out myself?”

If you’re a beginner to Python and you want to embark on this journey, then this post will guide you through your first steps. A common complaint I hear from complete beginners is that it’s pretty difficult to set up Python. How do we get everything started in the first place so that we can plug-and-play Data Science or Deep Learning code?

This post will guide you through in a step-by-step manner how to set up Python for your Data Science and Deep Learning projects. We will:

- Set up Anaconda and Jupyter Notebook
- Create Anaconda environments and install packages (code that others have written to make our lives tremendously easy) like tensorflow, keras, pandas, scikit-learn and matplotlib.

Once you’ve set up the above, you can build your first neural network to predict house prices in this tutorial here:

Build your first Neural Network to predict house prices with Keras

The main programming language we are going to use is called Python, which is the most common programming language used by Deep Learning practitioners.

The first step is to download Anaconda, which you can think of as a platform for you to use Python “out of the box”.

Visit this page: https://www.anaconda.com/distribution/ and scroll down to see this:

This tutorial is written specifically for Windows users, but the instructions for users of other Operating Systems are not all that different. Be sure to click on “Windows” as your Operating System (or whatever OS that you are on) to make sure that you are downloading the correct version.

This tutorial will be using Python 3, so click the green Download button under “Python 3.7 version”. A pop up should appear for you to click “Save” into whatever directory you wish.

Once it has finished downloading, just go through the setup step by step as follows:

Click Next

Click “I Agree”

Click Next

Choose a destination folder and click Next

Click Install with the default options, and wait for a few moments as Anaconda installs

Click Skip as we will not be using Microsoft VSCode in our tutorials

Click Finish, and the installation is done!

Once the installation is done, go to your Start Menu and you should see some newly installed software:

You should see this on your start menu

Click on Anaconda Navigator, which is a one-stop hub to navigate the apps we need. You should see a front page like this:

Anaconda Navigator Home Screen

Click on ‘Launch’ under Jupyter Notebook, which is the second panel on my screen above. Jupyter Notebook allows us to run Python code interactively on the web browser, and it’s where we will be writing most of our code.

A browser window should open up with your directory listing. I’m going to create a folder on my Desktop called “Intuitive Deep Learning Tutorial”. If you navigate to the folder, your browser should look something like this:

Navigating to a folder called Intuitive Deep Learning Tutorial on my Desktop

On the top right, click on New and select “Python 3”:

Click on New and select Python 3

A new browser window should pop up like this.

Browser window pop-up

Congratulations — you’ve created your first Jupyter notebook! Now it’s time to write some code. Jupyter notebooks allow us to write snippets of code and then run those snippets without running the full program. This helps us perhaps look at any intermediate output from our program.

To begin, let’s write code that will display some words when we run it. This function is called *print*. Copy and paste the code below into the grey box on your Jupyter notebook:

```
print("Hello World!")
```

Your notebook should look like this:

Entering in code into our Jupyter Notebook

Now, press Alt-Enter on your keyboard to run that snippet of code:

Press Alt-Enter to run that snippet of code

You can see that Jupyter notebook has displayed the words “Hello World!” on the display panel below the code snippet! The number 1 has also filled in the square brackets, meaning that this is the first code snippet that we’ve run thus far. This will help us to track the order in which we have run our code snippets.

Instead of Alt-Enter, note that you can also click Run when the code snippet is highlighted:

Click Run on the panel

If you wish to create new grey blocks to write more snippets of code, you can do so under Insert.

Jupyter Notebook also allows you to write normal text instead of code. Click on the drop-down menu that currently says “Code” and select “Markdown”:

Now, our grey box that is tagged as markdown will not have square brackets beside it. If you write some text in this grey box now and press Alt-Enter, the text will render it as plain text like this:

If we write text in our grey box tagged as markdown, pressing Alt-Enter will render it as plain text.

There are some other features that you can explore. But now we’ve got Jupyter notebook set up for us to start writing some code!

Now we’ve got our coding platform set up. But are we going to write Deep Learning code from scratch? That seems like an extremely difficult thing to do!

The good news is that many others have written code and made it available to us! With the contribution of others’ code, we can play around with Deep Learning models at a very high level without having to worry about implementing all of it from scratch. This makes it extremely easy for us to get started with coding Deep Learning models.

For this tutorial, we will be downloading five packages that Deep Learning practitioners commonly use:

- Set up Anaconda and Jupyter Notebook
- Create Anaconda environments and install packages (code that others have written to make our lives tremendously easy) like tensorflow, keras, pandas, scikit-learn and matplotlib.

The first thing we will do is to create a Python environment. An environment is like an isolated working copy of Python, so that whatever you do in your environment (such as installing new packages) will not affect other environments. It’s good practice to create an environment for your projects.

Click on Environments on the left panel and you should see a screen like this:

Anaconda environments

Click on the button “Create” at the bottom of the list. A pop-up like this should appear:

A pop-up like this should appear.

Name your environment and select Python 3.7 and then click Create. This might take a few moments.

Once that is done, your screen should look something like this:

Notice that we have created an environment ‘intuitive-deep-learning’. We can see what packages we have installed in this environment and their respective versions.

Now let’s install some packages we need into our environment!

The first two packages we will install are called Tensorflow and Keras, which help us plug-and-play code for Deep Learning.

On Anaconda Navigator, click on the drop down menu where it currently says “Installed” and select “Not Installed”:

A whole list of packages that you have not installed will appear like this:

Search for “tensorflow”, and click the checkbox for both “keras” and “tensorflow”. Then, click “Apply” on the bottom right of your screen:

A pop up should appear like this:

Click Apply and wait for a few moments. Once that’s done, we will have Keras and Tensorflow installed in our environment!

Using the same method, let’s install the packages ‘pandas’, ‘scikit-learn’ and ‘matplotlib’. These are common packages that data scientists use to process the data as well as to visualize nice graphs in Jupyter notebook.

This is what you should see on your Anaconda Navigator for each of the packages.

**Pandas:**

Installing pandas into your environment

**Scikit-learn:**

Installing scikit-learn into your environment

**Matplotlib:**

Installing matplotlib into your environment

Once it’s done, go back to “Home” on the left panel of Anaconda Navigator. You should see a screen like this, where it says “Applications on intuitive-deep-learning” at the top:

Now, we have to install Jupyter notebook in this environment. So click the green button “Install” under the Jupyter notebook logo. It will take a few moments (again). Once it’s done installing, the Jupyter notebook panel should look like this:

Click on Launch, and the Jupyter notebook app should open.

Create a notebook and type in these five snippets of code and click Alt-Enter. This code tells the notebook that we will be using the five packages that you installed with Anaconda Navigator earlier in the tutorial.

```
import tensorflow as tf
import keras
import pandas
import sklearn
import matplotlib
```

If there are no errors, then congratulations — you’ve got everything installed correctly:

A sign that everything works!

If you have had any trouble with any of the steps above, please feel free to comment below and I’ll help you out!

*Originally published by ** Joseph Lee Wei En** at *

===================================================================

Thanks for reading :heart: If you liked this post, share it with all of your programming buddies! Follow me on **Facebook** | **Twitter**

☞ A Complete Machine Learning Project Walk-Through in Python

☞ Machine Learning In Node.js With TensorFlow.js

☞ An A-Z of useful Python tricks

☞ Top 10 Algorithms for Machine Learning Newbies

☞ Automated Machine Learning on the Cloud in Python

☞ Introduction to PyTorch and Machine Learning

☞ Python Tutorial for Beginners (2019) - Learn Python for Machine Learning and Web Development

☞ Machine Learning A-Z™: Hands-On Python & R In Data Science

☞ Python for Data Science and Machine Learning Bootcamp

☞ Data Science, Deep Learning, & Machine Learning with Python