Tanmay Upadhyay - Portfolio

Home

Tanmay Upadhyay

Data Analyst | Data Scientist

Portfolio

Resume

Contact

I want to help others to make the right decisions, data-driven decisions

professional data-bender!

Skills

Machine Learning

I can build custom machine learning models using traditional algorithms and deep learning frameworks to automate processes and make accurate predictions based on your data.

Frameworks - PyTorch, TensorFlow, Scikit Learn

Data Engineering

With my expertise in data warehousing, ETL (Extract, Transform, Load) processes, and cloud computing technologies, I can help you build a robust and scalable data infrastructure. Few technologies I work with - Airflow, Snowflake, dbt etc.

Statistical Analysis

I can perform statistical analysis to validate hypotheses, test assumptions, and identify correlations in your data. Whether you need a regression analysis, a hypothesis test, or a time series analysis, I can help you make sense of your data and make informed decisions.

Data Analysis

Utilizing Python, SQL, and Spark, I analyze your data to uncover valuable insights such as patterns, correlations, and anomalies. My actionable recommendations empower you to make data-driven decisions and achieve your business goals.

Visualization & Dashboards

I can create interactive and engaging dashboards and visualizations using Tableau, Plotly and Matplotlib, that can help you communicate complex data insights in a simple and intuitive way.

Data Mining

I can apply data mining techniques to extract hidden insights from large datasets. PCA, T-SNE, LDA - are some of the techniques I can use to identify patterns, reduce dimensionality, and uncover hidden relationships in your data.

What i've done so far

Portfolio

Demystifying LA City Employee Payroll

North - American Avian Biodiversity Analysis

Spotify Recommendation System

Portfolio

World Development Indicator Analysis: Conducted cluster analysis of countries based on various development indicators using K-Neighbors and Hierarchical Clustering techniques. Utilized internal validation techniques to report cluster quality. Also derived insights based on various indicator combinations and their respective clusters.

source code

Financial Flow Type Classification: Deployed a KNN Classification Algorithm on python to classify foreign direct investments as either financial aid or non-aid (loans) with an accuracy of 95%. Created a dashboard that consists of all the analysis & insights, including interactive visualizations showing various trends in Investments throughout 10 years.

source code

dashboard

Network Analysis of Human Activity: Extracted time series features for various human activities and study its conservation when mapping time series into a network graph using HVG (Horizontal Visibility Algorithm). Established clear distinction between chaotic series and random series based on topological properties like degree distribution and mean path length.

source code

House Price Data Exploration in R: A Complete Exploratory Data Analysis (EDA) in R covering univariate and multivariate analysis of 10+ variables. Drawn correlation between various variables along with confirming normality, linearity, and homoscedasticity over a wide range of data. The Project walkthrough for the same is available on Medium.

source code

medium