Source Code Written By: Siddik Ayyappa
This is the source code of the work for Assignment-2, Statistical Methods in AI, 5th Semester, IIITH, '22. The assignment helps us to understand and thereby implement various machine learning techniques based on statistics, from scratch using numpy, pandas, frameworks.
- The First Question Involves the implementation of K-Nearest Neighbours classification algorithm. Done in Full.
- The Second question requires us to work out facial recognition with the help of dimensionality reduction. The technique of PCA (scratch) has been used on 672 Images of the IIIT-CFW dataset. dataset. The observations have been made in the python notebook, and soon a report will be put into the repository. eigenfaces. Done in full.
- The Third Question requires us to implement the logistic regression algorithm from scratch, on a dataset and plot the decision boundary. Done in full.
- The Fourth Question requires us to implement the vanilla linear regression algorithm from scratch on a dataset, plot the boundary, and then implement Ridge and Lasso Regression Techniques on the same data, understand the influence of the regularisation parameters on the accuracy of the model. Done in full.
- All the questions except for the 2nd one have been programmed object orientedly.
- All the questions give identical results when compared to the algorithms, implemented using frameworks like
sklearn
- I look forward to implement all of them in indivdual python files rather than notebooks.
- I would like to implement various other methods, like using different solvers.