Sergey Petrakov's repositories
Economics_Projects
Here you can observe some projects that I have made during my undergraduate studies at the Faculty of Economics of Lomonosov Moscow State University
Tensor_applications_to_RecSys
Here you can find project materials on term of tensor applications to recomendation systems. Mainly it is related to tensor decompositions, such as Tucker and CPD.
Uncertainty-Estimation-Transformers-in-Regression
Skoltech StatML laboratory project: Uncertainty Estimtimation for Trensformer models in regression task
CV
My CV
Deep-Learning-in-Asset-Pricing
https://arxiv.org/abs/1805.01104
Deep-Portfolio-Theory
Autoencoder framework for portfolio selection (paper published by J. B. Heaton, N. G. Polson, J. H. Witte.)
deep_learning_2018-19
Официальный репозиторий курса Deep Learning (2018-2019) от Deep Learning School при ФПМИ МФТИ
Diploma
Here is my diploma
fairseq
Facebook AI Research Sequence-to-Sequence Toolkit written in Python.
FamaFrench
Implementation of 5-factor Fama French Model
fse
this is my first Skoltech course repository
GENRE
Autoregressive Entity Retrieval
Hackathons
I am glad to share information about participation in hackathons, their materials and the results achieved
hyperopt
Distributed Asynchronous Hyperparameter Optimization in Python
kdnet.pytorch
implementation "Escape from Cells: Deep Kd-Networks for The Recognition of 3D Point Cloud Models" in pytorch
KILT
Library for Knowledge Intensive Language Tasks
Master_thesis
This repository contains part of my Master thesis "Uncertainty Enhances Knowledge Base Question Answering" at Skoltech that covers single based and MC-Dropout based metrics (ensemble consider output sequence as a possible class).
mGENRE_MEL
This repository contains files and materials related to multilingual entity linking task (MEL), especially basing on the mGENRE model since it is SOTA model. We consider MEL as a part of big knowledge base question answering (KBQA) that is called information retrieval part. Within this part we retrieve entities. Basing on them we can make queries to knowledge base. Thus, we obtain KBQA system
mlcourse.ai
Open Machine Learning Course
nla2021
The repository for Skolech NLA2021
Selective_Predictions_for_KBQA_and_El
This repository contains information about Selective prediction experiments of my master thesis for end-to-end KBQA model based on T5 and multilingual Entity Linking model based on mGENRE
Text_Classification
Text Classification Algorithms: A Survey