Completed Data Science projects throughout Yandex Practicum "Specialist in Data Science" training course
Link to visual portfolio - my portfolio
| Project name | Description | Libraries used |
|---|---|---|
| Sentiment analysis: Toxic comments identification | Building a system that identifies toxic comments and sends them to be edited/viewed. Building a model that classifies comments into positive and negative. | pandas numpy seaborn scipy datetime nltk xgboost lightgbm |
| Personal data encryption | Developing an encryption algorithm so that after data transformation, the accuracy of Machine Learning models will not change as without transformation. Building a system (ML model) on encrypted data. Creating a custom Linear Regresssion model. | pandas numpy seaborn scipy random |
| Forecasting taxi orders | Buidling a system for predicting the number of taxi orders for the next hour to attract more drivers during peak periods. Build a model that predicts number of taxi orders in next hour. | pandas numpy seaborn sklearn catboost xgboost lightgbm statsmodels |
| Optimization of gold mining company production | Optimizing gold mining production to avoid launching enterprises with unprofitable characteristics. Build a machine learning model to predict the recovery rate of gold from gold ore | pandas numpy matplotlib seaborn scipy sklearn |
| Pattern identification that determine success of the game | Analyzing open-source game data to identify patterns/metrics that determine the success of a game. Proposing recommendations for future actions. | pandas, numpy, matplotlib, seaborn, scipy |
| System to determine car value | Building a system to determine car value quickly using Gradient Boosting Machine Learning models | pandas numpy sklearn xgboost catboost lightgbm matplotlib seaborn scipy |
| Analyzing borrowers’ risk of defaulting | Evaluating the borrower's risk of defaulting. Finding possible causes of loan defaulting in bank customers data. Proposing a preferable client for the bank to take a loan. | pandas,numpy,matplotlib,seaborn,nltk,missingno |