toghrulutd Goto Github PK

-amazon-review-analysis-web-scraping-aspect-extraction-and-sentiment-analysis

This repository includes web-scraping, cleaning, preprocessing of Amazon customer reviews and aspect extraction and sentiment classification of reviews with spaCy language models and Deep Learning models.

abstract-sequence-classification-nlp

The goal of this project is to build a hybrid NLP model (RNNs, BERT) to classify medical abstract sentences into the role they play (e.g. objective, methods, results, etc) to enable researchers to effectively skim through the literature and dive deeper when necessary.

adtracking-fraud-detection

Data engineering and machine learning pipeline for detecting ad fraud using TalkingData AdTracking dataset. The project includes data exploration, preprocessing, feature engineering, model training, and evaluation.

amex-credit-card-default-prediction

In this project, I have used industry scale customer data to predict credit card default probability using Ensemble learning methods. I have carried out extensive data preprocessing, feature engineering and selection, tuned model parameters and implemented various model explainability approaches to explain errors.

awesome-mlops

A curated list of references for MLOps

awesome-production-machine-learning

A curated list of awesome open source libraries to deploy, monitor, version and scale your machine learning

car-appraisal-app

This is a Flask web app that predicts the price of used car based on its make, model, bodystyle, mileage, year, and location. The repository also covers web scraping the required data, model building and tuning.

ensemble-regression-for-price-estimation

In this project, I have used Gradient Boosting Regressor to predict used car prices from Poland's used car market. The project covers data preprocessing, handling categorical variables, feature transformation/engineering, and application and tuning of linear and ensemble learning methods, model explainability and residual analysis.

food101-with-sharpness-aware-minimization

The objective of this project is to implement transfer learning and recent finetuning techniques proposed by two research papers. In this project, we use sharpness aware minimization, and MC dropout techniques to build an accurate image classification model with least training and beat the performance of original paper.

instacart-basket-analysis

Instacart: Customer Purchase Behaviour Analysis and Product Recommendations

loan-approval-classification-with-imbalanced-data

This project includes tasks such as applying DART ensemble learning, model tuning, data preprocessing, feature engineering, model explainability and residual analysis

microbusiness-density-forecasting-with-n-beats-

The goal of this project is to develop an ensemble of Neural Basis Expansion networks trained on U.S. county-level data to predict monthly micro-business density in a given area for 8-month horizon.

movie-recommender-systems

This repository contains implementations of various recommender systems for the Movielens dataset, including matrix factorization with TensorFlow and Spark, Bayesian inference, restricted Boltzmann machines, and deep learning recommenders. The code is provided in Jupyter notebooks and Python scripts, along with notes on these topics.

toghrulutd.github.io