The time-series forecasting is a vital area that motivates continuous investigate areas of intrigued for different applications. A critical step for the time-series forecasting is the right determination of the number of past observations (lags). This paper investigates the forecasting accuracy based on the selection of an appropriate time-lag value by applying a comparative study between three methods. These methods include a statistical approach using auto correlation function, a well-known machine learning technique namely Long Short-Term Memory (LSTM) along with a heuristic algorithm to optimize the choosing of time-lag value, and a parallel implementation of LSTM that dynamically choose the best prediction based on the optimal time-lag value. The methods were applied to an experimental data set, which consists of five meteorological parameters and aerosol particle number concentration. The performance metrics were: Mean Absolute Error (MAE), Root Mean Square Error (RMSE), Mean Absolute Percentage Error (MAPE) and R-squared. The investigation demonstrated that the proposed LSTM model with heuristic algorithm is the superior method in identifying the best time-lag value.

Time-series prediction is an important area that inspires numerous research disciplines for various applications, including air quality databases. Developing a robust and accurate model for time-series data becomes a challenging task, because it involves training different models and optimization. In this paper, we proposed and tested three machine learning techniques—recurrent neural networks (RNN), heuristic algorithm and ensemble learning—to develop a predictive model for estimating atmospheric particle number concentrations in the form of a time-series database. Here, the RNN included three variants—Long-Short Term Memory, Gated Recurrent Network, and Bi-directional Recurrent Neural Network—with various configurations. A Genetic Algorithm (GA) was then used to find the optimal time-lag in order to enhance the model’s performance. The optimized models were used to construct a stacked ensemble model as well as to perform the final prediction. The results demonstrated that the time-lag value can be optimized by using the heuristic algorithm; consequently, this improved the model prediction accuracy. Further improvement can be achieved by using ensemble learning that combines several models for better performance and more accurate predictions.

In this paper, we present deep learning models for classifying ancient-coins using transfer learning applied to a custom dataset curated from a specialized collection. The dataset was meticulously developed through rigorous preparation and quality screening; the retained fine-grained classes are imbalanced. We initially performed binary classification before progressing to separate fine-grained assessments of DenseNet121, EfficientNetB0, EfficientNetV2S, and ResNet50. In confirmatory experiments using a physical-coin-aware 70/15/15 split and five training seeds, the coarse CITY-versus-NABATAEAN task remained approximately perfectly separable. Fine-grained performance was lower: ResNet50 achieved 68.28 ± 2.45% CITY accuracy, whereas DenseNet121 achieved 72.89 ± 2.79% NABATAEAN accuracy. These results show that near-perfect coarse classification does not by itself imply reliable fine-grained attribution. This analysis underscores the importance of meticulous dataset preparation and strategic model selection in optimizing performance. A matched DenseNet121 ablation further showed that ImageNet initialization substantially improved the difficult fine-grained tasks compared with random initialization. The proposed framework supports digital documentation and analysis of ancient coin collections, while the current single-collection evaluation still limits claims of cross-collection generalization.

Machine-learning intrusion detection on multi-attack Internet-of-Things (IoT) datasets is often reported through aggregate metrics that can hide near-failure on rare attack classes, and through single hand-picked feature subsets whose stability is seldom examined. Using the real TON_IoT Network dataset (211,043 flows, 10 classes, with man-in-the-middle (MITM) traffic at 0.5% prevalence), this study contributes a transparent empirical mapping rather than a new detector. A strong full-feature Random Forest reaches 0.969 accuracy and 0.945 macro-averaged F1 (macro-F1) yet recalls only 0.776 of the rare MITM class, and a mutual-information filter raises aggregate macro-F1 to 0.948 while leaving worst-class recall unchanged at 0.776. A self-coded non-dominated sorting genetic algorithm (NSGA-II) is then used to map the trade-off between macro-F1 and worst-class recall as the selected subset varies, alongside a matched-size random-subset control, standard imbalance remedies, and two additional classifiers. Three findings are robust. First, imbalance remedies on the full feature set—class weighting and, especially, the synthetic minority over-sampling technique (SMOTE)—raise MITM recall to 0.818 and 0.904, matching or exceeding the wrapper, so feature selection is not the most effective tool for the rare class. Second, re-adding a single destination-port feature inflates ransomware recall by 0.094 and cross-site-scripting (XSS) recall by 0.066 while doing nothing for MITM, showing that several apparently detected classes are separable by an identifier artifact whereas the genuinely hard class is not. Third, selected subsets are unstable (Nogueira index 0.21) despite reproducible objective values. The wrapper’s own rare-class advantage is small and does not reach significance under a cross-validation-appropriate corrected test. The two transferable findings replicate on the NSL-KDD benchmark. We argue that IoT intrusion-detection studies should report the aggregate–worst-class frontier, an identifier ablation, a selection-stability index, and cross-validation-appropriate significance tests.

Machine-learning models that predict short-horizon cellular throughput underpin adaptive video streaming, cloud gaming, and edge orchestration. On the widely used Raca et al. 5G drive-test dataset, such models can report low error. We show that, on this dataset, these headline numbers can be optimistic and, more importantly, that aggregate metrics conceal where and how the models fail. Using the saturated file-download sessions, where throughput reflects deliverable capacity rather than application idling, we (i) quantify the optimism of random-row evaluation relative to a deployment-realistic leave-sessions-out (LOSO) protocol, negligible at a 1 s horizon but reaching ∼32% at 8 s, and show that a trivial persistence baseline significantly beats the learned models at the two shortest horizons; (ii) show by ablation that most reported accuracy is carried by autoregressive throughput features rather than radio/context indicators and that the indicator-only mapping generalizes poorly across sessions; (iii) find, contrary to the intuition that handovers dominate error, that absolute error is lower around transitions and concentrated in the high-throughput regime (Gini 0.71), but that, in scale-invariant terms, the error is largest at low throughput and that transitions carry only a modest, noisy under-prediction tendency (event-aligned trough ≈$-6$ Mbps) that nearly vanishes under linear-space training (≈−1 Mbps), i.e., largely a log-space retransformation artifact; and (iv) find, using paired per-session bootstrap tests, that lightweight transition-context features give no significant global gain, a (mostly) null result. All headline numbers carry per-session bootstrap confidence intervals; the self-lag finding further replicates on an independent HSDPA drive-test dataset from Australia. We recommend leave-sessions-out evaluation, a persistence baseline, self-lag ablation, and scale-invariant, bias-aware reporting as a minimal honest protocol.