Shweta Mishra

Responsible and motivated Data Science Master’s student with prior professional experience, ready to apply education in the workplace. Offers excellent technical abilities with statistical analysis and Machine Learning applications, ability to handle challenging work, and excellent time management skills.

Skin Cancer Diagnosis with DL and CV | Deep Learning, Transfer Learning, CNNs, Flask 01/2023 - 03/2023

• Detected skin cancer with deep learning and computer vision, achieving 89.5% accuracy on HAM10000 dataset. Improved lesion classification by 4.7% using VGG16 and custom CNNs. Devised Flask based diagnostic web app.

Market Basket Analysis & Recommendation System | Apriori, Collaborative Filtering, K-Means, KNN 10/2022 - 11/2023

• Optimized recommendation and marketing strategies using algorithms like Apriori, Item-Item and User- User Collaborative Filtering, and customer segmentation via clustering using KMeans on a 200K user Instacart dataset.

News Text Summarization | NLP, BART, ROUGE Score, Bleu Score 08/2022 - 09/2023

• Transformed news text summarization with extractive and abstractive approaches, resulting in remarkable ROUGE-1: 0.8 and BLEU score of 0.85. Utilized GloVe embeddings, seq2seq architecture, and pre-trained T5 model.

Hack Harvard 2022 | Assembly AI, Streamlit Major League Hacking (MLH) 10/2022 - 10/2022

• Created ZMatter Open-Source app in Hackathon (36 hours) to provide insights (Text, Summary, and Gist) from educational content of TikTok videos using AssemblyAI API. Invited for a Web3 grant from PI Network.

Community Member: Rewriting The Code (RTC), Women Who Code, The Data Club, GHC '23, ODSC, Hacktoberfest, MLH, HackHarvard