Dmitrii's repositories
Reinforcement-learning-in-Trading
Reinforcement learning in trading and algorithmic trading is a fairly frequent example of the application of RL in practice.
Airflow-ETL-Example
This repository contains 2 DAGs for working with databases.
Application-of-generative-models-for-prediction-of-drug-molecular-compounds
Данная статья является победителем Международного Салона Образования. Текст публикации доступен по ссылке: https://top-technologies.ru/ru/article/view?id=39067
Customer-Segmentation
In this task, the task of segmentation of clients (clustering by the K-Means method and Hierarchical clustering) was performed
EDA-With-R-Vaccine
This analytical repository shows the use of the R language in the field of EDA. Data on the success of vaccination against coronavirus infection were taken as data. Various graphs, histograms and graphs for countries of the world were constructed, showing the success of individual countries in vaccinating people.
FlappyBird-PyTorch-1.11-Reinforcement-Learning-DDQN
In this project, the agent learns to play FlappyBird based on the DDQN model.
Fraud-Detection
It is important that credit card companies are able to recognize fraudulent credit card transactions so that customers are not charged for items that they did not purchase.
Gender-Age-Ethnicity-Recognition
This analytical project has three objectives: 1) Gender recognition based on a person's photo; 2) Recognition of ethnicity by photo; 3) Age recognition based on a person's photo
Java-ATM-PetProj
This program implements the interface of a bank ATM. By logging in with a username and PIN, the user should be able to make deposits, withdraw funds and transfer funds between accounts. The user should also be able to list transactions for each account, so this program can be something in between the functionality of an ATM and an online banking system. The goal of this project is to master the basic features of the Java language. The project is a homework assignment on the subject of the Java Programming Language.
News-Classification-Fake-True
This analytical task is primarily aimed at classifying news (fake or real).
Recommendation-system-of-films
In this task, a decision was made to build 3 models of recommendation systems based on the MovieLens Small dataset.
REST-API-With-Golang
Another RESTFUL API using Golang, which is written using the Gin framework, which allows you to use HTTP requests to fill, extract or delete data.
Speech-Emotion-Recognition
This project, using a neural network, allows you to determine emotions in human speech.
Twitter-Sentiment-Analysis
This analytical project allows you to determine the mood of users by text. The sentimentality 140 dataset is used as a dataset. It contains 1,600,000 tweets extracted using the twitter api. The streets have been labeled (0 = negative, 4 = positive), and they can be used to determine moods. The following works were carried out: research data analysis, text preprocessing (removal of stop words, punctuation marks, lemmatization, TF-IDF, tokienization). In the end, two models were built: the basic model (a model based on the naive Bayes model - Rein et al. (2003), Tackling the Poor Assumptions of Naive Bayes Text Classifiers). The top 10 tokens that characterize the text as negative or optimistic were extracted, but they turned out to be non-offensive. The second model was based on Bert (87% accuracy rate).
Web-ChatBot
Chat bot based on a neural network using flask
Brain_Hemorrhage_Classifier_OctConvResnet
This notebook code is written based upon the interest in Kaggle Competition (https://www.kaggle.com/c/rsna-intracranial-hemorrhage-detection ) using newly launched OctConv (https://ai.facebook.com/blog/octconv-a-flexible-efficient-alternative-to-standard-convolution/) by Facebook
ChayannFamali
Config files for my GitHub profile.
keras
Deep Learning for humans