arXiv subjects

Samiul Islam

Publications and source records attributed to Samiul Islam.

3 recordsLinked to original sources

arXiv2026

Green-NAS: A Global-Scale Multi-Objective Neural Architecture Search for Robust and Efficient Edge-Native Weather Forecasting

We introduce Green-NAS, a multi-objective NAS (neural architecture search) framework designed for low-resource environments using weather forecasting as a case study. By adhering to 'Green AI' principles, the framework explicitly minimizes computational energy costs and carbon footprints, prioritizing sustainable deployment over raw computational scale. The Green-NAS architecture search method is optimized for both model accuracy and efficiency to find lightweight models with high accuracy and very few model parameters; this is accomplished through an optimization process that simultaneously optimizes multiple objectives. Our best-performing model, Green-NAS-A, achieved an RMSE of 0.0988 (i.e., within 1.4% of our manually tuned baseline) using only 153k model parameters, which is 239 times fewer than other globally applied weather forecasting models, such as GraphCast. In addition, we also describe how the use of transfer learning will improve the weather forecasting accuracy by approximately 5.2%, in comparison to a naive approach of training a new model for each city, when there is limited historical weather data available for that city.

cs.LG↗

arXiv2025

Tunable Valley Polarization and Anomalous Hall Effect in Ferrovalley NbX2 and TaX2 (X = S, Se, Te): A First-Principles Study

Two-dimensional transition metal dichalcogenides lack inversion symmetry and have broken time-reversal symmetry due to the honeycomb structure and intrinsic ferromagnetism, which leads to their valley polarization. Here, we explored the electronic and magnetic properties of the novel ferrovalley materials 1H-NbS2, 1H-NbSe2, 1H-NbTe2, 1H-TaS2, 1H-TaSe2, and 1H-TaTe2 using first-principles calculations based on density functional theory. The materials are dynamically stable bipolar magnetic semiconductors. Among the magnetic semiconductors, NbSe2 showed the maximum Curie temperature of 176.25 K. For these materials, the ferromagnetic state was more favorable than the antiferromagnetic state, indicating robust ferrovalley characteristics. These ferrovalley materials showed a giant tunable valley polarization at K and K' points in the Brillouin zone without applying any external factors due to intrinsic exchange interactions of transition metal d-orbital electrons and spin-orbit coupling. TaTe2 exhibited an outstanding valley splitting of 541 meV. Reversing Bloch electrons' magnetic moment caused an alteration of valley polarization. Additionally, the application of uniaxial and biaxial strain led to the manipulation and variation of the bandgap and valley polarization. Berry curvature exhibited opposite signs and unequal magnitudes at K and K' points, which led to the anomalous valley Hall effect in these materials. NbS2, NbSe2, and NbTe2 exhibited Berry curvature at unstrained crystals, whereas Berry curvature appeared only in TaSe2 and TaTe2 with the application of strain. These ferrovalley materials exhibited distinct band gaps for spin-up and spin-down electrons, enabling the selective transport of spin-polarized electrons.

cond-mat.mtrl-sci↗

arXiv2023

Capturing Spectral and Long-term Contextual Information for Speech Emotion Recognition Using Deep Learning Techniques

Traditional approaches in speech emotion recognition, such as LSTM, CNN, RNN, SVM, and MLP, have limitations such as difficulty capturing long-term dependencies in sequential data, capturing the temporal dynamics, and struggling to capture complex patterns and relationships in multimodal data. This research addresses these shortcomings by proposing an ensemble model that combines Graph Convolutional Networks (GCN) for processing textual data and the HuBERT transformer for analyzing audio signals. We found that GCNs excel at capturing Long-term contextual dependencies and relationships within textual data by leveraging graph-based representations of text and thus detecting the contextual meaning and semantic relationships between words. On the other hand, HuBERT utilizes self-attention mechanisms to capture long-range dependencies, enabling the modeling of temporal dynamics present in speech and capturing subtle nuances and variations that contribute to emotion recognition. By combining GCN and HuBERT, our ensemble model can leverage the strengths of both approaches. This allows for the simultaneous analysis of multimodal data, and the fusion of these modalities enables the extraction of complementary information, enhancing the discriminative power of the emotion recognition system. The results indicate that the combined model can overcome the limitations of traditional methods, leading to enhanced accuracy in recognizing emotions from speech.

cs.SD↗