COPS IG AI/ML/Data Science Roadmap for Beginners
Hola Amigos freshers,
“Artificial Intelligence, Machine Learning and Data science”
All these words sound so fascinating and overwhelming. But to have the right start you need to first understand the differences between the three, explore them and discover your interest!
P.S. Go step by step. We have attached video links alongside.
Link: AI VS ML VS DL VS Data Science - YouTube
Inorder to understand in depth each of them, choose the one excites you the most or understand them and decide.
By the end of this section you need to choose one to begin
Artificial Intelligence
Artificial intelligence is a broad area of computer science that makes machines seem like they have human intelligence.
Example: If you have ever asked Alexa to order your food or browse Netflix movie suggestions, you are interacting with AI without realizing it!
Link : What Is Artificial Intelligence? | Artificial Intelligence (AI) In 10 Minutes | Edureka - YouTube
Machine Learning
Machine learning is a thing-labeler where you explain your task with examples instead of instructions. The concept behind Machine Learning is that you feed data to machines and let them learn on their own without any human intervention (in the process of learning).
Examples: Classifying whether the image is of cat or dog. (Stay tuned, you’ll be able to do it yourself very soon)
Link: Machine Learning Basics | What Is Machine Learning? | Introduction To Machine Learning | Simplilearn - YouTube
Data Science
Data science is a field focused on discovering actionable insights from large sets of raw data. Data scientists use different techniques to get answers incorporating computer science, analysis, statistics, and machine learning on massive datasets to establish solutions to problems.
Example: Determining who died and who survived in Titanic sinking using dataset, (You’ll do this very soon)
Link: Data Science In 5 Minutes | Data Science For Beginners | What Is Data Science? | Simplilearn - YouTube
We hope you have chosen one. Now let’s quickly jump to the road map.
Roadmap: Data Science
Step 1:
We highly recommend you to use Google Colab. However you can use Jupyter Notebook too.
Link: Google Colab Tutorial for Beginners | Using google Colab for machine learning and Deep learning - YouTube
Link to Google Colab: Welcome To Colaboratory - Colaboratory (google.com)
(Watch the video to learn basics on how to use Colab)
Step 2:
Learn the basics of Python to get started.
P.S. This is very important! Python is the most used language for ML/AI ( Although it can be done in other languages, we highly highly recommend you to use Python).
P.S. If you are already well versed with Python, feel free to skip this part.
Link: Python Tutorial for Beginners 2: Strings - Working with Textual Data - YouTube
Top Tip: 1. Watch ONLY video number 2-9.
2. Try out everything on Colab yourself.
Step 3:
For Data Science, python has some specific libraries that will help you to make all your tasks easier.
Do Learn each of them.
Pandas:
Short : Link: What is Pandas in Python - Introduction to Pandas for Data Science - Python - Jupyter Notebook - YouTube
Long: Link: Python Pandas Tutorial in Hindi - YouTube
(We recommend you to watch the longer one for better understand of pandas as it is extremely important library)
Numpy:
Link: Creating One Dimensional Array Using Numpy - In Hindi - YouTube
Watch the videos from number 4 to 8.
Matplotlib:
Link: [Hindi] Python Matplotlib Tutorial - Python Data Science and Big Data Tutorials - YouTube
Step 4:
Now since you
are done with understanding libraries. These libraries help you in not only working on the dataset but also understanding the dataset. Data Visualization plays a key role in Data science.
Link:
Data Visualization | Data Visualization Python | Intellipaat - YouTube
Exercise: In the above video they have used ‘Iris dataset’ as an example for visualizing the dataset. Your task is to use the famous ‘Titanic dataset’ for the same.
How to download? Go to the link Titanic - Machine Learning from Disaster | Kaggle Make an account on Kaggle. Accept to Join the Competition. Click below on ‘Download All’.
After trying it yourself. Watch the video below. Solution: Link: Python pandas—Visualization Exercises—Titanic - YouTube
P.S. A lot more can be done. This is an example.
Step 5:
In order to predict our result from a dataset. Machine Learning uses different learning algorithms to make the task efficient. These are present pre-built in scikit-learn library. However it is equally important to understand the math behind it.
There are many learning algorithms. Here we will learn two beginner friendly ones:
- Linear Regression
Link: Lecture 2.1 — Linear Regression With One Variable | Model Representation — Andrew Ng - YouTube
(Watch videos from 2.1 to 2.7 from the playlist & Do not skip any of them)
- Logistic Regression for Classification problems
Link: Lecture 6.1 — Logistic Regression | Classification — — [ Machine Learning | Andrew Ng] - YouTube (Watch the videos from 6.1 to 6.7 from the playlist & Do not skip any of them)
Step 6:
After understanding the theory behind the algorithms, it is equally important to be able to implement it for data science. Now it’s your time to code!
Implementation of Linear Regression:
Link: Machine Learning Tutorial Python - 3: Linear Regression Multiple Variables - YouTube
Implementation of Logistic Regression:
Link:Machine Learning Tutorial Python - 8: Logistic Regression (Binary Classification) - YouTube (DO NOT watch entire video.
Time stamp: 8:18 to 14:33)
Top Tip: While watching the video, try to implement yourself on Colab simultaneously. Remember you can’t learn it until you do it yourself.
Step 7:
Let’s get your hands dirty. Solve the ‘Titanic dataset challenge’ yourself and take your first step to become a data scientist!
Hint: Titanic dataset uses Logistic Regression.
Message: You learn only by competing in data science competitions. Titanic dataset is beginner friendly and solutions can be easily found online. You’ll be surprised to see how quickly and accurately you’ll be able to determine whether the person survived the sinking! IT’S LIKE MAGIC. Try it and share your score on ML discord channel.
Link to challenge:
Titanic - Machine Learning from Disaster | Kaggle
Help: Titanic Tutorial | Kaggle (This tutorial will guide you through the Challenge)
Step 8:
Don’t stop here and keep exploring.
Some useful resources:
1.Krish Naik YouTube Playlist. Best Explanation of topics Learning Data Science In 2022- Step By Step Plan - YouTube
2.StatQuest YouTube playlist
A Gentle Introduction to Machine Learning - YouTube
Do check out this playlist to learn Machine Learning Algorithms, Graphs and more math.
Machine Learning or Deep Learning
Introduction: Machine Learning is a very broad field that has various applications. We’ll start with Computer Vision (CV). It is often implemented with a very famous library called OpenCV (Open Source Computer Vision Library). One of the major uses of Open CV is detecting and working on images. We’ll learn more further.
P.S. It has lots of theory!
Step 1:
We highly recommend you to use Google Colab. However you can use Jupyter Notebook too.
Link: Google Colab Tutorial for Beginners | Using google Colab for machine learning and Deep learning - YouTube
Link to Google Colab: Welcome To Colaboratory - Colaboratory (google.com)
(Watch the video to learn basics on how to use Colab)
Step 2:
Learn the basics of Python to get started.
P.S. This is very important! Python is the most used language for ML/AI ( Although it can be done in other languages, we highly highly recommend you to use Python).
P.S. If you are already well versed with Python, feel free to skip this part.
Link: Python Tutorial for Beginners 2: Strings - Working with Textual Data - YouTube
Top Tip: 1. Watch ONLY video number 2-9.
2. Try out everything on Colab yourself.
Step 3:
Understand what you’re doing. Andrew NG course is the best course to understand the math behind ML. Machine Learning - Introduction | Coursera Machine Learning - Linear Algebra Review | Coursera
(Complete ONLY Week 1 & Week 3)
Step 4:
After understanding the theory behind the Linear Regression algorithm, it is equally important to be able to implement it for ML. Now it’s your time to code!
Implementation of Linear Regression:
Link: Machine Learning Tutorial Python - 3: Linear Regression Multiple Variables - YouTube
(Try all this code on Colab YOURSELF)
Step 5:
Before diving deep into math. It’s important for you to be acquainted with some important libraries required in ML.
Link: Python: Top 5 Machine Learning Libraries - YouTube
P.S Eventually you got to learn all these. But we can hold on for a while for this roadmap.
Step 6:
Get your hands dirty on some code. So that you don’t lose touch with coding while learning the new interesting stuff ahead.
Link: Scikit Learn Tutorial | Machine Learning with Python | Python for Data Science Training | Edureka - YouTube
P.S. It’s important to code right now so that you understand the importance of what you’ll be learning ahead.
Step 7:
Now we are jumping onto something very important in ML.The NEURAL NETWORKS!!
What is CNN?
A convolutional neural network (CNN) is a type of artificial neural network used in image recognition and processing that is specifically designed to process pixel data.
CNNs are powerful image processing, artificial intelligence (AI) that use deep learning to perform both generative and descriptive tasks, often using machine vision that includes image and video recognition, along with recommender systems and natural language processing (NLP).
Link: Convolutional Neural Networks (CNNs) explained - YouTube
Tutorial 20- Convolution Neural Network vs Human Brain - YouTube
Step 8:
Understand working behind CNNs
Link: Convolutional Neural Networks - Foundations of Convolutional Neural Networks | Coursera
(Complete ONLY Week 1)
Step 9:
Before we learn implementation for CNN. We need to be acquainted with another famous library of ML called Tensorflow.
First, understand what is tensorflow
Link:TensorFlow In 10 Minutes | TensorFlow Tutorial For Beginners | Deep Learning & TensorFlow | Edureka - YouTube
Understand each and every step of model training minutely
Link: TensorFlow Tutorial 2 - Tensor Basics - YouTube
(DO NOT SKIP THIS PART)
P.S Watch videos 2 to 5 ONLY
Step 10:
Implement it on solving real world problems.
We’ll be doing it on a very famous CIFAR 10 dataset.
Link: Image classification using CNN (CIFAR10 dataset) | Deep Learning Tutorial 24 (Tensorflow & Python) - YouTube
Also develop a habit of reading papers and Medium articles
Link: Deep Learning with CIFAR-10. Neural Networks are the programmable… | by Aarya Brahmane | Towards Data Science
Top tip: Open Incognito if they ask for login
Link for dataset:
CIFAR-10 - Object Recognition in Images | Kaggle
Step 11:
Now do it yourself. Let’s get your hands dirty. Solve the ‘DIgit Recognizer’ yourself and take your first step to become a ML engineer!! Try making your computer recognize which handwritten digit it is, simply by an input image of the numeric digit.
Message: You learn only by competing in ML competitions. MNIST is beginner friendly and solutions can be easily found online. You’ll be surprised to see how quickly and accurately you’ll be able to determine which digit is written! IT’S LIKE MAGIC. Try it and share your score on ML discord channel.
Link to challenge:
MNIST Handwritten Digit Recognition | Kaggle
(This tutorial will guide you through the Challenge)
Step 12:
Don’t stop here and keep exploring. Some useful resources.
1.Krish Naik YouTube Playlist. Best Explanation of topics Why Deep Learning Is Becoming So Popular?🔥🔥🔥🔥🔥🔥 - YouTube
2.Continue for other weeks Andrew NG Convolutional Neural Networks - Foundations of Convolutional Neural Networks | Coursera
Good Luck!