IGIgor Gavrilov

Igor Gavrilov

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Igor Gavrilov has a total of 4 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Hybrid SARIMA-recurrent neural network approach for road accident forecasting

    by Igor Gavrilov, Pavel Hrubes, Emil Pelikan

    This study investigates the forecasting of daily road traffic accidents using time series models, through a controlled comparative evaluation of baseline SARIMA and Hybrid SARIMA-LTSM frameworks. We employ a Seasonal Autoregressive Integrated Moving Average (SARIMA) model to capture linear and seasonal patterns in accident data from 2010 to 2023. To address potential non-linearities, we extend the approach by modelling SARIMA residuals using Recurrent Neural Networks (RNNs), including bidirectional RNN (BiRNN) and Long Short-Term Memory (LSTM) variants, to develop a hybrid SARIMA-RNN framework. The hybrid model integrates SARIMA forecasts with RNN-based residual predictions as a methodological extension whose empirical performance is rigorously evaluated against the baseline. Within residual-domain model selection, the unidirectional LSTM achieved the lowest RMSE (0.184) and MAPE (1.0966%) among tested architectures, confirming its effectiveness in capturing temporal dependencies in the log-residual space. A hybrid SARIMA-LSTM model was also evaluated to enhance residual prediction; however, the one-step-ahead rolling-origin strategy with a 7-day sliding window did not improve accuracy relative to the baseline SARIMA, yielding only minor increases in RMSE and MAPE.

  • A Neural Network-Based Framework for Identifying Hazardous Road Segments Using Geometrical Road Attributes

    by Igor Gavrilov, Pavel Hrubes, Emil Pelikan

    Accurate identification of hazardous road segments is essential for improving road safety and prioritising preventive interventions. This study proposes a neural network-based framework for predicting hazardous road segments by leveraging detailed geometrical properties of roads, including lane number, lane width, and curvature. The proposed framework was applied to the D1 highway in the Czech Republic, using geospatial data and reported road accidents from 2024. The road geometry was divided into 10 m sectors, from which sequences of 300 m were generated to form input for a Long Short-Term Memory (LSTM) network. The model was trained using a weighted binary cross-entropy loss function to address class imbalance, where hazardous segments represented only 29.2 % of samples. The trained model demonstrated strong predictive performance, achieving a weighted $\mathbf{F 1}$ score of 0.895 and a macro F1 score of 0.906, which effectively distinguished between hazardous and non-hazardous highway segments. The results demonstrate that integrating detailed road attributes within an LSTM-based framework can effectively capture spatial dependencies and non-linear relationships associated with accident occurrence. The framework provides a scalable approach for identifying high-risk segments across various road types, offering valuable support for data-driven decision-making in transportation safety management.