This is a complete machine learning course for computer science students, Electronics and communication students, and any other graduate students, This course deals with both supervised and unsupervised models present in machine learning. This course gives a clear understanding of machine learning to those who want to change their domain towards machine learning.
The entire course is arranged in a sequence. those who follow this sequence will get a clear understanding of mathematics and machine learning models that are used in the industry. With this course you can face the interview questions easily. and the implementation of all the models is also explained. This gives a better practical approach to machine learning.
The first section is from the first to the thirteenth video. this section gives the mathematics basics required to understand machine learning. This includes the concepts of slope of a line, the slope of a function at a given point, the derivative of a function, finding the maxima and minima point on a function, gradient descent, how gradient descent helps in finding the minimum point in a given function, linear regression model understanding, stochastic gradient descent, batch gradient descent, and multiple linear regression.
The second section is from the fourteenth video to the fifty videos. This section gives a good understanding of data preprocessing, data cleaning, and model efficiency calculation.
This includes data cleaning, data preprocessing, min-max scaling, normalization, standardization, one hot enoding, stemming, stop words, a bag of words, tfidf, word2vec, training data, validation data, testing data, overfitting models, underfitting models, types of supervised models, calculating accuracy in regression and classification, K fold cross-validation, confusion matrix, the problem with imbalanced data, and ROC curve.
The third section is about the Naive Bayes model. which includes Bayes theorem, intuition on naive Bayes classification, Laplace smoothing, hyperparameter underfitting and overfitting, log probability, multinomial naive Bayes, and Gaussian naive Bayes implementation in python.
The fourth section is K nearest neighbors model. this includes euclidean distance, manhattan distance, Minkowski distance, decision surface, hyperparameter bias-variance tradeoff in KNN, the curse of dimensionality, problem with KNN, and the implementation of KNN in Python.
The fifth section is the Logistic regression model. this includes algebra basic for logistic regression, sigmoid function, the problem with distance measure in logic regression, loss function logistic regression, regularization in logistic regression, hyperparameter bias-variance tradeoff, solving optimization problem logistic regression, and the implementation in python.
The next section is the Support vector machine model, this includes Lagrange multiplier, karush khun tuker examples, functional margin, geometric margin, primal and dual problem, optimization problem SVM, solving optimization problem SVM, use of transforming data to high dimension, kernel function, hyperparameter bias-variance tradeoff, soft margin SVM, and the code in python.
The next section is the Decision tree model. this includes graphical intuition on decision trees, entropy, information gain, constructing decision tree, Gini impurity, information gain on numerical features, hyperparameter bias-variance tradeoff, decision tree regression, Graphviz for constructing decision tree and visualization, and implementation in python.
The next section is ensemble techniques.
This include.
1) bagging
2) Boosting
3) stacking
The random forest comes under the bagging technique.
The gradient boosting comes under the boosting technique.
Finally stacking techniques.
The next section is unsupervised models.
This include
1) K means clustering
2) Hierarchical clustering
K means include k mean++ initialization, knee or elbow method for right k, silhouette method for right k, k means clustering failure cases, and implementation in python.
The hierarchical model includes agglomeration, divisive, dendrogram, similarity measures, and agglomeration clustering code in python.