Sonali Agarkar

May 2021- April 2022
KEY CORSEWORKS IN MSc
▪ Completed dissertation project, ‘An Explainable AI Based

Approach for Skin Cancer Classification’. Analysed the data and visualised the insights. Compared performance of conventional machine learning models, SVM and Random Forest with CNN and Resnet50. Provided visualisations of explanations for model output of Resnet50 and CNN with LIME and SHAP. Tools used are Keras Tensorflow, SciKit learn, Motplotlib and Seaborn.

▪  Created Entity Relationship model for the given case. Performed database management tasks: creating tables, inserting values, extracting data using SQL queries.

▪  Performed analysis on aviation dataset. Created interactive visualisations using Tableau and Seaborn, Matplotlib. Machine learning is algorithms are applied applied using PySpark.

▪  Applied Deep learning methods ANN and CNN on retinal images to predict the presence of cataract. Tools used are Keras, Tensorflow, Scikit learn.

▪  Analysed gene expression using Statistical Methods. Applied Non-Linear Regression and computed model parameters, Log likelihood, AIC, and BIC values. The analysis is done using R on R-studio.

▪  Used machine learning techniques to predict possible loan defaulters. After preprocessing the dataset, Random Forest, Logistic Regression, Adaboost and SVM algorithms are applied, and performances are compared.

2022
COVENTRY UNIVERSITY, UK
MSc in Data Science and Computational Intelligence Distinction

2003
AMRAVATI UNIVERSITY, INDIA
Bachelor of Engineering in Electronics and Telecommunication
First Class