Макаров Илья Андреевич

Instagram Hashtag Prediction Using Deep Neural Networks

2021 · CHAPTER · en

*Реализация соц. сети Instagram запрещена на территории России по основаниям осуществления экстремистской деятельности. Instagram is one of the most popular photos sharing services. For more convenient content search people use hashtags (#nature, #love, etc.) in posts with photos. The author’s aim is to make hashtag prediction possible and convenient for users. The paper provides a reader with a detailed theoretical overview of Multi-Label Image Classification, Knowledge Distillation, and an overview of ResNet architecture. Next, the author proposes improvements on ResNet architecture allowing the model to boost quality and converge faster. Finally, the model type Self-Improving-Modified-Resnet (SIMR) is presented. Their main feature is the additional bottleneck block used as the tool incorporating benefits from a combination of self-training and knowledge distillation.

JONNEE: Joint Network Nodes and Edges Embedding

2021 · ARTICLE · en

Recently, graph embedding models significantly improved the quality of graph machine learning tasks, such as node classification and link prediction. In this work, we propose a model called JONNEE (JOint Network Nodes and Edges Embedding), which learns node and edge embeddings under self-supervision via joint constraints in a given graph and its edge-to-vertex dual representation as a Line graph. The model uses two graph autoencoders with additional structural feature engineering and several regularization techniques to train for an adjacency matrix reconstruction task in an unsupervised setting. Experimental results show that our model performs on par with state-of-the-art undirected attribute graph embedding models and requires less number of epochs to achieve the same quality due to Line graph self-supervision under a unified embedding framework.

Depth Inpainting via Vision Transformer

2021 · CHAPTER · en

Learning Loss for Active Learning in Depth Reconstruction Problem

2021 · CHAPTER · en

Style-transfer Autoencoder for Efficient Deep Voice Conversion

2021 · CHAPTER · en

Network Embedding for Cluster Analysis

2021 · CHAPTER · en

Graph visualization is an effective and efficient way to discover complex inter-connections between elements within the nested structure of data. To accomplish this type of representation machine learning algorithms use a technique called graph embedding and node embedding in particular. However, in this paper, we will compare well-known techniques to yet largely under-explored setting of graph embedding named community embedding: embedding individual communities instead of individual nodes. This type of embedding can be especially useful in graph visualization and community detection tasks. Despite the fact that graph embedding and clustering tasks are separate, a good solution to the first one tends to have a correlation with the solution of the second problem and may have a positive impact if knowledge is transferred.

Deep Reinforcement Learning with DQN vs. PPO in VizDoom

2021 · CHAPTER · en

Prediction of New Itinerary Markets for Airlines via Network Embedding

2020 · CHAPTER · en

A large number of methods are being developed in the deep reinforcement learning area recently, but the scope of their application is limited. The number of environments does not always allow for a comprehensive assessment of a new agent training algorithm. The main purpose of this article is to present another environment for Match-3 game that could be expanded, which would have a connection with the real business. The results for the most popular deep reinforcement learning algorithms are presented as a baseline.

GSM: Inductive Learning on Dynamic Graph Embeddings

2020 · CHAPTER · en

In this paper, we study the problem of learning graph embeddings for dynamic networks and the ability to generalize to unseen nodes called inductive learning. Firstly, we overview the state-of-the-art methods and techniques for constructing graph embeddings and learning algorithms for both transductive and inductive approaches. Secondly, we propose an improved GSM based on GraphSAGE algorithm and set up the experiments on datasets CORA, Reddit, and HSEcite, which is collected from Scopus citation database across the authors with affiliation to NRU HSE in 2011–2017. The results show that our three-layer model with attention-based aggregation function, added normalization layers, regularization (dropout) outperforms suggested by the respective authors’ GraphSAGE models with mean, LSTM, and pool aggregation functions, thus giving more insight into possible ways to improve inducting learning model based on GraphSAGE model.

Collaborator Recommender System

2020 · CHAPTER · en

Nowadays, a lot of scientists’ works aim to improve the quality of people’s life but it could be quite complicated without building a successful collaboration. Productive partnerships can increase research efficiency in many cases and make a huge impact on society. However, today there is no clear way to find such collaborators. In this paper, we propose a recommender system for the scientists from the Higher School of Economics university to help them find co-authors for their prospective studies.