Duration**2 Months**

Mode of Training**Online **

Level**Advanced**

Machine learning (ML) is a field of study in artificial intelligence concerned with the development and study of statistical algorithms that can learn from data and generalize to unseen data and thus perform tasks without explicit instructions.

This course will introduce the learner to applied machine learning, focusing more on the techniques and methods than on the statistics behind these methods. The course will start with a discussion of how machine learning is different than descriptive statistics, and introduce the scikit learn toolkit through a tutorial. The issue of dimensionality of data will be discussed, and the task of clustering data, as well as evaluating those clusters, will be tackled. Supervised approaches for creating predictive models will be described, and learners will be able to apply the scikit learn predictive modelling methods while understanding process issues related to data generalizability. The course will end with a look at more advanced techniques, such as building ensembles, and practical limitations of predictive models. Develop an understanding of AI & ML and its components with the PG Programme for working professionals in Artificial Intelligence & Machine Learning, Deep Learning.

- What Is Machine Learning?
- Supervised Versus Unsupervised Learning
- Regression Versus Classification Problems Assessing Model

- Supervised Versus Unsupervised Learning
- Introduction to Matrices
- Vector spaces, including dimensions, Euclidean spaces, closure properties and axioms
- Eigenvalues and Eigenvectors, including how to find Eigenvalues and the corresponding Eigenvectors

- Linear Regression
- Simple Linear Regression
- Estimating the Coefficients
- Assessing the Coefficient Estimates
- Squared and Adjusted R SquaredV
- M SE and RMSE
- Estimating the Regression Coefficients
- OLS Assumptions
- Multicollinearity
- Feature Selection
- Gradient Discent

- Homoscedasticity and Heteroscedasticity of error terms
- Residual Analysis
- Q-Q Plot
- Cook's distance and Shapiro-Wilk Test
- Identifying the line of best fit
- Other Considerations in the Regression Model
- Qualitative Predictors
- Interaction Terms
- Non-linear Transformations of the Predictors

- Why Polynomial Regression
- Creating polynomial linear regression
- evaluating the metrics

- Lasso Regularization
- Ridge Regularization
- ElasticNet Regularization

- Logistic regression
- An Overview of Classification
- Difference Between Regression and classification Models
- Why Not Linear Regression?
- Logistic Regression:
- The Logistic Model
- Estimating the Regression Coefficients and Making Pr edictions
- Logit and Sigmoid functions
- Setting the threshold and understanding decision boundary
- Logistic Regression for >2 Response Classes
- Evaluation Metrics for Classification Models:
- Confusion Matrix
- Accuracy and Error rate
- TPR and FPR
- Precision and Recall, F1 Score
- AUC â€“ ROC
- Kappa Score
- Principle of Naive Bayes Classifier
- Bayes Theorem
- Terminology in Naive Bayes
- Posterior probability
- Prior probability of class
- Likelihood
- Types of Naive Bayes Classifier
- Multinomial Naive Bayes
- Bernoulli Naive Bayes and Gaussian Naive Bayes

- Decision Trees
- Decision Trees (Rule Based Learning)
- Basic Terminology in Decision Tree
- Root Node and Terminal Node
- Regression Trees and Classification Trees
- Trees Versus Linear Models
- Advantages and Disadvantages of Trees
- Gini Index
- Overfitting and Pruning
- Stopping Criteria
- Accuracy Estimation using Decision Trees

- Cross-Validation
- The Validation Set Approach Leave-One-Out Cross-Validation
- k -Fold Cross-Validation
- Bias-Variance Trade-Offfor k-Fold Cross-Validation
- What is Ensemble Learning?
- What is Bootstrap Aggregation Classifiers and how does it work?
- What is it and how does it work?
- Variable selection using Random Forest
- is it and how does it work?
- Hyper parameter and Pro's and Con's

- Cross-Validation
- The Validation Set Approach Leave-One-Out Cross-Validation
- k -Fold Cross-Validation
- Bias-Variance Trade-Offfor k-Fold Cross-Validation
- What is Ensemble Learning?
- What is Bootstrap Aggregation Classifiers and how does it work?
- What is it and how does it work?
- Variable selection using Random Forest
- is it and how does it work?
- Hyper parameter and Pro's and Con's

- K-Nearest Neighbor Algorithm
- Eager Vs Lazy learners
- How does the KNN algorithm work?
- How do you decide the number of neighbors in KNN?
- Curse of Dimensionality
- Pros and Cons of KNN
- How to improve KNN performance
- The Maximal Margin Classifier
- HyperPlane
- Support Vector Classifiers and Support Vector Machines
- Hard and Soft Margin Classification
- Classification with Non-linear Decision Boundaries
- Kernel Trick
- Polynomial and Radial
- Tuning Hyper parameters for SVM
- Gamma, Cost and Epsilon
- SVMs with More than Two Classes

- Why Unsupervised Learning
- How it Different from Supervised Learning
- The Challenges of Unsupervised Learning
- Principal Components Analysis
- Introduction to Dimensionality Reduction and it's necessity
- What Are Principal Components?
- Demonstration of 2D PCAand 3D PCA
- EigenValues, EigenVectors and Orthogonality
- Transforming Eigen values into a new data set
- Proportion of variance explained in PCA
##### Case Study: A Case Study on PCA using Python.

##### K-Means Clustering

- Centroids and Medoids
- Deciding optimal value of 'k' using Elbow Method
- Linkage Methods

- Divisive and Agglomerative Clustering
- Dendrograms and their interpretation
- Applications of Clustering
- Practical Issues in Clustering
- What are recommendation engines?
- How does a recommendation engine work?
- Data collection
- Data storage
- Filtering the data
- Content based filtering
- Collaborative filtering
- Cold start problem
- Matrix factorization
- Building a recommendation engine using matrix factorization

- Introduction to Neural Networks
- Introduction to Perceptron & History of Neural networks
- Activation functions a)Sigmoid b) Relu c)Softmax d)Leaky Relu e)Tanh
- Gradient Descent
- Learning Rate and tuning
- Optimization functions
- Introduction to Tensorflow
- Introduction to keras
- Back propagation and chain rule
- Fully connected layer
- Cross entropy
- Weight Initialization
- Regularization
- TensorFlow 2.0
- Introducing Google Colab
- Tensorflow basic syntax
- Tensorflow Graphs
- Tensorboard
- Neural Network for Regression
- Neural Network for Classification

- Introduction to NLP
- Preprocessing, NLP Tokenization, stop words, normalization, Stemming and lemmatization
- Preprocessing in NLPBag of words, TF-IDF as features
- Language model probabilistic models, n-gram model and channel model
- Hands on NLTK
- Word2vec
- Golve
- POS Tagger
- Named Entity Recognition(NER)
- POS with NLTK
- TF-IDF with NLTK
- Introdcution to sequential models
- Introduction to RNN
- Intro to LSTM
- LSTM forward pass
- LSTM backprop through time
- Hands on keras LSTM
- Sentiment Analysis
- Sentence generation
- Machine translation
- Advanced LSTM structures
- Keras- machine translation
- ChatBot