A Mathematical and Programming Course on Machine Learning

Name: A Mathematical and Programming Course on Machine Learning
Rating: 3.6 (7 reviews)

In this course the mathematical concepts of machine learning will be taught to learners, with python in Google Colab

Created byRituraj Dixit

Last updated 2/2023

English

What you'll learn

In depth knowledge of mathematics behind building ML models
How to prepare data for feeding into models
In depth analysis of support vector machines and their kernels
Concepts of Ensemble methods in machine learning
Building Recommendation system by using concepts of machine learning
Building Recommendation System
Implementation of CNN models
Implementation of Fashion MNIST
Recurrent Neural network
Quiz at the end of each Section to test the concepts you have learned
Natural Language Processing
Active Learning
Implementation of Cost Estimation functions using TensorFlow from scratch

Course content

16 sections • 128 lectures • 34h 33m total length

Detail Overview of Course10:28
This video will give students information about the outcomes of the course and the detail course structure which is composed of nine sections. They will get an idea about how to work on the course and what are the topics covered in this course
How to connect Google CoLab with .csv file7:31
This video will tell learner how to link their .csv file which is stored in google drive with google colab IDE, a very useful video connecting data with IDE.
Basics of Python : Part1(Basic Constructs, IF Else)9:11
This Lecture is the first lecture of Section 1 of Basics of python module which give concepts of programming for students who are new to python programming. we have given basic introduction along with conditional statements in python using google colab in this lecture
Basics of Python : Part2(Loops in python)9:11
This lecture is second in the series of basics of python, in this lecture concepts of loops are introduced to students along with examples.
Basics of Python : Part3(List)24:07
This Lecture will give detail implementation of list in python
Basics of Python : Part4(Dictionary and Tuples)15:21
Important concepts of Dictionary is being introduced in this lecture along with tuples.
Basics of Python : Part5(Functions)17:07
In this lecture students will get an idea about the functions in python along with practical exercise
Concepts of Numpy in Google Colab41:54
Numpy is a very important package to understand carefully for building an efficient machine learning algorithm, so it is highly recommended to understand numpy in detail
Concepts of Pandas part19:22
Concepts of Pandas part212:49
Exploratory data analysis using test case-part114:43
The car_sample.csv file is provided to all of you have to build the model from scratch as being told in this lecture it consist of three parts, EDA, Data Cleaning , Model building and comparison with Random Forest Regressor
Data Cleaning using Test case part-237:18
This lecture will describe how to build the first model using the concept of data exploration and data cleaning, after this we reduce the dimension of the data by using ffeature engineering and chosen best model to build one of the possible best model

Different Types of Learning23:47
Bayesian Learning and NB classifier25:02
Naïve Bayes Classifier6:59
Bayesian Network14:25
Implementation of Naïve Bayesian Classifier in Python13:35
Practical Implementation of Naïve Bayes Classifier20:18
Concept of different type of Regularization9:23
Regularization Theory : Practical Implementation of Ridge and Lasso Regression23:06
Principal Component Analysis and LDA : part112:57
Principal Component Analysis and LDA : part213:05
Implementation of Linear and Quadratic Discriminant Analysis20:02
Building PCA from Scratch using Python16:42
KNN Algorithm using Python12:53
Quiz onMachine learning

Concept of Linear and Logistic Regression18:39
Implementation of linear regression from scratch16:22
An example of Ordinary Least Square7:47
Introduction to Multiple Linear Regression24:45
Importance of significance test in multiple linear regression11:42
Practical Implementation of Multiple Linear regression20:42
Maximum Likelihood estimation importance in Logistic Regression23:59
Maximum Likelihood Estimation is use to determine the pdf value of a random value, which means the probability of random variable to belong a particular class or distribution
Logistic Regression and analysis10:30
Implementation of Logistic Regression in python18:37
in this lecture we will implement a logistic regression classifier, which is used to check whether the customers using the coupon or not and their dependency on card and amount expenditure
Loan Prediction algorithm using Logistic Regression32:05
In this test cases we have used Loan prediction data set for training the model, along with it i have also provided a test case which you can use to predict model values.
The CHI square is used to predict the significance of independent variable with target variable both are categorical.
To check the over all significance we have used Variance Inflation Factor for checking importance of variable
Feature Engineering on continuous data : House Prediction22:03
In this lecture we perform various important task on the training data like handling missing values dividing data and follow divide and conquer approach, then we applied feature engineering on continuous data
Feature Engineering on Categorical Variable : House Prediction12:51
This lecture is in continuation with the above lecture here we will see how to choose important categorical features from a pool of features
Quiz on Linear and Logistic Regression

Overview of Section 4 Decision Tree and its implementation7:00
Introduction to Decision tree and ID323:41
Introduction to ID3 and its concepts21:30
Concept of Entropy, Overfitting and Information Gain13:31
Practical Implementation of ID3and their limitations16:10
C4.5 Decision and its advancement over ID3 Decision tree6:50
Building a Decision tree with data using python from scratch22:51
Practical Implementation of CART Algorithm using Python27:19
This algorithm will discuss the cart algorithm implemented on a small data set to identify fruits
CHAID Algorithm and its importance in Data Analytics22:58
CHAID is a very important decision tree which is based on the concept of finding chi square to check the significance of independent variable with respect to dependent variable. The selection of the feature or attribute depend upon the highest chi square value.
Implementation of CHAID using Python10:34

Introduction to Recommendation System (RecSys)6:16
Memory based Collaborative Filtering (CF)18:42
Matrix Decomposition based Collaborative Filtering15:08
Item based Recommendation System8:16
Movie Recommendation using Content Based Filtering32:44
Collaborative Filtering based Movie Recommendation16:56
Hybrid Collaborative filtering36:25
In this lecture students will learn how to combine Content based rating and Item Based Collaborative rating.
Implementation of clustering in Recommendation System8:03
Implementation of Machine Learning in Recommendation System8:49
Quiz of Recommendation System

Introduction to Support Vector Machines13:16
The Constrained Optimization Problem20:48
The dual formulation concept of SVM part116:16
The dual formulation concept of SVM part215:32
The concept of soft margin classifier in SVM is explained in detail in this lecture along with the concept of dual form of Lagrange expression which has some condition which are very important in optimizing the solutions
Maximum margin with the introduction of noise8:50
After completing this lecture students will understand the concept of support vector machine more clearly, they also came to know how to maximize the margin separation between two classes
Non Linear SVM Classifier15:07
Implementation of Linear and Non Linear Support Vector Classifier26:37
Implementation of SVR(Support Vector Regressor9:47
Multi SVM and their applications7:20
Case Study: Implementation of Support Vector Classifier12:57
Case Study : Implementation of Support Vector Regressor6:55
Support Vector Machine Quiz

Introduction of Tensor flow in Neural Network and Deep Networks13:04
Data representation in the form of Tensors35:04
Concept of Artificial Neural Network8:32
The Perceptron4:38
Practical implementation of Perceptron4:42
Gradient Descent Algorithm using python21:48
This lecture will give students information about gradient descent algorithm which is a very important mechanism for parameter tuning.
Stochastic gradient Descent Algorithm using Python13:17
Mini Batch Gradient Descent Algorithm12:59
Practical implementation of MLP17:40
Back Propagation Algorithm14:15
Batch Normalization13:10
Building a Classifier for Fashion MNIST dataset37:46
Implementation of Linear regression using Tensor flow on Housing data15:24
Practical implementation of Convolution Neural Network12:30
A brief introduction of Computer vision using openCV21:10
Quiz on ML and Deep Learning

Introduction to the concepts of Unsupervised Learning and clustering1:31
The Mathematics of K-Means Clustering9:53
The math of k means part 24:52
Practical Implementation of K-Means Clutering17:57
Concept of Silhouette Coefficient to measure quality of Clusters13:06
The Mathematics of Hierarchal Agglomerative Clustering17:44
Implementation of Hierarchal Agglomerative Clustering using PCA20:29
Fuzzy C Means Algorithm-FAANY8:20
Fuzzy C Means Algorithm : Practical Implementation22:58
Practical implementation of FCM algorithm using Python
Mean Shift Clustering11:05
DBSCAN- Density Based Spatial Clustering16:27
Gaussian Mixed Model with Expectation Maximizing Clustering30:41
GMM Clustering - Practical Implemtation11:13
This lecture gives a practical aspect of how clusters are viewed as Gaussian Normal Distributions. In this lecture we try to separate overlapping clusters
Quiz based on concept of Clustering

Introduction to Random Forest25:37
Voting Classifier and Bagging Classifier16:24
The Concepts of Voting classifier both soft voting using predict_proba() method and hard voting classifier have been explained in this lecture using data. The bagging classifier is also being implemented to give students an idea about both types of classifiers
Random Forest Classifier10:46
The random forest is an ensemble of lot of decision trees breast cancer data set has been used to implement this classifier
Random Forest Regressor and its comparison with Linear Regression model -part35:44
Ada Boost Implementation : Test Case 17:38
Gradient Boosting Regression Trees9:39
This lecture deals with building a Gradient Boosting Regression Tree, a very powerful model which work on the base estimator which is Decision Tree Regressor. Then we have introduced stage predict() method which work on the concept of early stopping

Requirements

No Prerequisites, only will to learn

Description

This course of "A Comprehensive Course on Machine Learning using python" is a very comprehensive and unique course in itself. Machine Learning is a revolution now days but we cannot master machine learning without getting the mathematical insight, and this course is designed for the same. Our course starts from very basic to advance concepts of machine learning. We have divided the course into different modules which start from the introduction of python its programming basic and important programming constructs which are extensively used in ML programming.

The mathematics involved in Machine learning is normally being not discussed and being left out in , but in our course we have put lot of emphasis in mathematical formulation of algorithms used in ML. We have also designed modules of pandas, sklearn, scipy, seaborn and matplotlib for gearing the students with all important tools which are needed in dealing with data and building the model. The machine learning module focuses on the mathematical derivation on white board through video lectures because we believe that white box view of every concept is very important for becoming an efficient ML expert.

In Machine Learning the cost estimation function also called loss functions are very important to understand and in our course we have explained Cross Categorical Entropy, Sparse Categorical Cross Entropy, and other important cost functions using TensorFlow.

Concepts like gradient descent algorithm, Restricted Boltzmann Algorithm, Perceptron, Multiple Layer Perceptron, Support Vector Machine, Radial Basis Function , Naïve Bayes Classifier, Ensemble Methods, recommendation system and many more are being implemented with examples using Google Colab.

Further I wish best of luck to learners for their sincere efforts in advance…

Use of various components of statistics in analyzing data
Graphical representation of data to get deep insight of the patterns
Mathematical analysis of algorithms to remove the black box view
Practical implementation of all important ML Algorithms
Building various models from scratch using advance algorithms
Understanding the use of ML in research
Quiz at the end of each section

Who this course is for:

undergraduates, graduates who want to learn python and machine learning along with their mathematical concepts

A Mathematical and Programming Course on Machine Learning

What you'll learn

Explore related topics

Course content

A mathematical approach to Machine learning : An Overview12 lectures • 3hr 29min

Statistics using Python2 lectures • 16min

Machine Learning using Python13 lectures • 3hr 32min

Linear Regression and Logistic Regression using Python12 lectures • 3hr 40min

Decision Trees and its implementation in Python10 lectures • 2hr 52min

Recommendation System using Python9 lectures • 2hr 31min

Support Vector Machines and its types11 lectures • 2hr 33min

Neural Network and Deep Neural Network15 lectures • 4hr 6min

Un Supervised Learning and Concepts of Clustering13 lectures • 3hr 6min

Ensemble methods and Random Forest6 lectures • 1hr 16min

Requirements

Description

Who this course is for: