Deep Learning Specialization Course 3

Deep Learning Specialization Course 3

It's been a long since I started learning this specialization course but in between, I feel overwhelmed about the kind of content available for me to learn in the field of machine learning and keep on browsing the other things losing touch with...

It's been a long since I started learning this specialization course but in between, I feel overwhelmed about the kind of content available for me to learn in the field of machine learning and keep on browsing the other things losing touch with this one. This time I have decided to complete the certification and then only move forward with practicing other courses.

Have you felt similar challenges while planning to switch careers in this really lucrative field? If yes, do post in the comments. We together can definitely figure out a way to have a blast in the field of data science.

Without any delay, let's dive into learning how to structure a machine learning project.

Why ML Strategy

Let’s start with an example; We have built a cat classifier and achieved a 90% accuracy. Now to improve the classifier, we might try several things such as gathering more data, apply L2 regularization, try to drop out some units, or trying bigger or smaller networks. But if we move forward in the wrong direction there is a possibility that we end up spending a lot of time with no fruitful output. Wouldn’t it be nice if we had quick and effective ways to identify which of these or other ideas are worth pursuing? Hence in this course, we will learn several strategies and ways of analyzing ML problems that will point us in the most promising direction.

Orthogonalization

One of the challenges of machine learning systems is that there are several things that we can try while training the model. One of the important steps is to tune hyperparameters. The process of clearly identifying what to tune to achieve a particular effect is known as Orthogonalization.

For supervised learning systems to do well, we usually need to make sure that four things hold true.

  1. The model should work well on the training set. It should be equal to human-level performance. We can use a bigger network or switch to a better algorithm to ensure better performance.
  2. Fit development set well on the cost function. We can utilize regularization or bigger training set to ensure good performance on the dev set.
  3. Fit the test set well on the cost function. If the model does not do well on the test set but does well on the dev set then we will have to find a bigger dev set as it might have been over-tuned.
  4. Perform well in the real world. If the model does not perform well in the real world then we might have to change the dev-test set distribution or the cost function we have selected is not performing as expected.

We will see each point in detail.

data-science machine-learning deep-learning deeplearningai

What is Geek Coin

What is GeekCash, Geek Token

Best Visual Studio Code Themes of 2021

Bootstrap 5 Tutorial - Bootstrap 5 Crash Course for Beginners

Nest.JS Tutorial for Beginners

Hello Vue 3: A First Look at Vue 3 and the Composition API

Most popular Data Science and Machine Learning courses — July 2020

Most popular Data Science and Machine Learning courses — August 2020. This list was last updated in August 2020 — and will be updated regularly so as to keep it relevant

Machine Learning Vs Deep Learning: Difference Between Machine Learning and Deep Learning

This article will simply explain the concept which will help you understand the difference between Machine Learning and Deep Learning. 

Hire Machine Learning Engineer | Offshore Machine Learning Experts

We are a Machine Learning Services provider offering custom AI solutions, Machine Learning as a service & deep learning solutions. Hire Machine Learning experts & build AI Chatbots, Neural networks, etc. 16+ yrs & 2500+ clients.

How To Build A Data Science Career In 2021

In Conversation With Dr Suman Sanyal, NIIT University,he shares his insights on how universities can contribute to this highly promising sector and what aspirants can do to build a successful data science career.

PyTorch for Deep Learning | Data Science | Machine Learning | Python

PyTorch for Deep Learning | Data Science | Machine Learning | Python. PyTorch is a library in Python which provides tools to build deep learning models. What python does for programming PyTorch does for deep learning. Python is a very flexible language for programming and just like python, the PyTorch library provides flexible tools for deep learning.