Pallavi Bhimte
Data Scientist | Data Analyst
Helping people to make data-driven decisions.
Built a job-hunting web-based data application with flask that uses natural language processing model trained with pre-trained FastText embedding vectors.
List of things I can do for you
Machine
Learning
Natural Language
Processing
Statistical
Modelling
Data
Visualization
Data
Analysis
Story
Boards
Projects
DigiEmploy - NLP Web App
Quality Prediction for Learning Resources
Developed Random Forest and SVM classification models with improved accuracy(better than baseline performance) that classifies the customer engagement based on several learning resource features such as visual quality, description length, etc.
Quality Prediction for Learning Resources
Bias Eradication using AIF360
Eradicated gender bias from all stages of machine learning pipeline (pre-processing, in-processing, and post-processing) with improved AdaBoost classification accuracy by implementing reweighing algorithm from AI Fairness 360 toolkit for customer’s credibility prediction based on their credit card details
Australia’s Population Growth - Story Board
Story Telling with Open Data:
Published a storyboard on RPubs with plotted visualizations from eight open-source data sets that narrates a story and identifies the main factor responsible for population growth in Australia by interpretation of yearly birth rate, death rate, fertility rate, age distribution, etc.Australia’s Population Growth - Story Board
Stroke Prediction
ML models to predict whether a person is likely to experience a stroke or not based on various factors(age, marriage, hypertension, smoker, etc.). Applied appropriate feature selection, model fitting, and hyperparameter fine-tuning to identify the best predictive model for model comparison
Sunspots Forecasting and Prediction
Time Series Analysis: Examined monthly mean sunspot numbers over 60 years of data(1960-2020) with the interpretation of seasonality, trend, behavior, intervention point, and change of variance.
Reported prediction results for the next two years and generated forecasts for the next 10 years from the best ARIMA model obtained.Sunspots Forecasting and Prediction
Fertility Prediction
Regression Analysis: Executed three predictive modeling techniques including linear, logistic, and multiple regression, and predicted the fertility rate using the best model. Checked adequacy of each model using various assumptions such as non-constant variance, linearity, normal distribution of residuals, autocorrelation, and ANOVA test.
International Airlines Passenger Seat Distribution
Executed main data wrangling steps such as subsetting the data, changing wide to long format, factoring and generating a new variable, scanning for missing values and outliers, and applying appropriate data transformation techniques including box-cox, square-root, cube, log10, and log transformation
International Airlines Passenger Seat Distribution
Vulnerability Calculator
Proposed a solution to indicate how vulnerable the person is towards getting infected by covid based on people’s demographics, health status, and occupation. Designed and developed an interactive browser-based application in RStudio(Shiny Apps) to help hospitals, government, public, and healthcare organizations in their work to identify and plan strategies to reduce the impact of COVID-19 by assigning a risk value to the user.
Recommendations
Pallavi was involved in one of our data science projects at Go1. She brought excellent data analysis and data science expertise, along with curiosity. Besides that, I really enjoyed her teamwork attitude and her help to organise the project tasks to achieve a goal as a team. She is also eager to learn new things which is very valuable. All the best Pallavi and I hope we cross paths again in the future.
Pallavi has been an excellent team player where she was responsible for coming up with some strategies with regard to server configurations and laying down the network policies for the organisation. She was part of a team where she contributed to other activities and participated actively in various projects. She is a hard worker and talented in her field of expertise. and it was indeed a delight to work with her.