Advisor Resume Example

All data science projects done below are written in either Jupiter Notebook or Google Colab.

Project 1: Recommendation system

• Use the Amazon dataset contains ratings of different electronic products to build a recommendation system to recommend products to customers based on their previous ratings for other products.
• Rank based recommendation system, Collaborative Filtering based recommendation system, Model based Collaborative Filtering using Singular Value Decomposition.

Project 2: Applying dimensionality reduction on the auto-mpg dataset and doing bank customer segmentation

• Explore the auto-mpg dataset by applying dimensionality reduction techniques and visualize the data in lower dimensions to extract insights. Segment the bank customers to help the bank upgrading the service delivery model and ensure that customers queries are resolved faster
• PCA, Clustering, tSNE.

Project 3: Analyzing the marketing campaigns to help CMO devise the next strategy

• Analyze marketing data and address some important business problems/questions related to customer demographics, product preferences, channel performances, etc.
• Exploratory Data Analysis, Data Visualization, Statistics .

Project 4: Street view housing number digit recognition

• Use one of the most popular image recognition datasets SVHN dataset to perform number recognition.
• Tuned hyperparameters including number of hidden layers, dropout layers, activation functions, maxpooling, flatten, padding. Selected optimizer Adam from tensorflow.keras.
• CNN, ANN.

Project 5: Predicting the eligibility of bank customers for a loan and forecasting the stock prices

• Predict the eligibility of the bank customers for a loan by using classification algorithms. And forecast the stock prices for the next 24 months using time series.
• Decision Trees, Random Forest, Time Series.

Project 6: Predicting the Boston house prices and eligibility of bank customers for a loan

• Predict the house prices based on the features of the locality using regression techniques. And predict the loan eligibility of bank customers using classification algorithms to help the bank prevent potential losses, and focus more on eligible customers.
• Linear Regression, Logistic Regression, kNN.