Vadim Sedelnikov

Projects: System for the GOE steel production: • Preprocessing and combining data for all cycle of industrial production • Quality prediction & Recommendation of optimal parameters Server Capacity planning and forecasting: • Forecasting time series • Predicting based on business metrics (baskets, purchases, users) Driving Style Identification (taxi drivers) with Unsupervised Learning • Anomaly detection • Driver clustering Text Analyzer • Pattern detection and recognition

Models and Methods: Logistic, Linear Regressions, Random Forest, Decision Tree, Boosting, Bagging, K-means, DBSCAN, T-SNE, SVM, PCA, One-hot-encoding, OOB-method, Tf-idf, Bag-of-words, L1-L2 regularization, Scaling and others Association Rules (Apriori) Python (Scikit-learn, Scipy, Pandas, Seaborn, Matplotlib, NumPy, XGBoost, iGraph) H2O Google Geocoder API github.com/VadimSSS

Thesis: Adaptive algorithms of information processing in the automatic detection and recognition of objects The award of "The grant of Moscow"

Master's thesis: The analysis of random signals using Kohonen neural network. The award of "Grant of Soros Foundation"