Training projects translated into English
| Project name | Description | Used libraries |
|---|---|---|
| Customer outflow forecasting | Building a model that predicts whether the user will leave | Pandas, Matplotlib, NumPy, Sklearn, Seaborn, CatBoost, Phik |
| Age determination by photo | Building a model that could predict a person’s age | Pandas, Matplotlib, NumPy, Keras |
| Definition of toxic comments | ML-model of machine learning allows to classify comments on positive and negative, in order to send them to moderation | Pandas, NumPy, ntlk, Sklearn, tqdm |
| Taxi order forecasting | Construction of the forecast system for the number of taxi orders for the next hour | Pandas, Matplotlib, NumPy, Sklearn, CatBoost, Statsmodels |
| Car cost determination | For the service of determining the market value of the car at the request of the client, a model predicting the price of the car is implemented | Pandas, Matplotlib, NumPy, Sklearn, Seaborn, CatBoost |
| Personal data protection | Development of data encryption method. Construction of machine learning model on encrypted data | Pandas, NumPy, Sklearn |
| Investigation of the gold refining process | In order to optimize production, a prototype ML-model for predicting the recovery rate of gold from gold-containing ore | Pandas, Matplotlib, NumPy, Sklearn, Seaborn, SciPy |
| Location Selection for Oil Well | Finding a profitable oil-producing region by simulating machine learning | Pandas, Matplotlib, NumPy, Sklearn |
| Customer outflow forecasting | On the basis of data from the bank determine the client who may leave | Pandas, Matplotlib, Sklearn |
| Tariff recommendation | Mobile operator "Megaline" wants to build a system capable to analyze the behavior of customers and offer users a new tariff. Task: to construct a model for the task of classification, which will choose a suitable tariff | Pandas, Sklearn |
| Research into the success of computer games | Open Data Analysis and Search for a Potentially Popular Product that Allows the Customer to Plan Advertising Campaigns | Pandas, Matplotlib, NumPy, Seaborn, SciPy |
| Definition of advantageous tariff for the telecom company | A preliminary analysis of the tariffs of the telecom company on a sample of 500 customers was conducted and the most profitable tariff was identified for subsequent adjustment of the advertising budget | Pandas, Matplotlib, NumPy, Seaborn, SciPy |
| Research on advertisements for the sale of apartments | Using the data of the service Yandex.Real Estate, determine the market value of real estate and typical parameters of apartments | Pandas, Matplotlib, NumPy, Seaborn |
| Borrower reliability study | On the basis of statistics on customers' ability to investigate whether the client’s marital status and number of children affect the fact of repayment of the loan on time | Pandas |
| City music | On real data Yandex.Music with the help of the Pandas library and its ability to check the data and compare the behavior and preferences of users of the two capitals - Moscow and Saint Petersburg | Pandas |