An optimized Indian-General-Elections-Based social science data prediction using multiscale dense nested parallel MobileNetV3 mantis search attention network
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
Forecasting an election outcome is a challenging exercise because of the constantly varying and numerous political, social, and demographic factors. Unlike electoral data, no work explores the local structure in large-scale data since traditional approaches fail to capture complex patterns involving large sets of data, especially from diverse areas such as India. This research aims to present an improved, large-scale scale, and efficient prediction model for Indian General Elections through a Multiscale Dense Nested Parallel MobileNetV3 Mantis Search Attention Network, 3MDNPV3-SAN. The model is developed to overcome the shortcomings common to other techniques and incorporates several approaches. For data preprocessing, the novel Anisotropic Gaussian Filtering with Directionally Truncated First Derivative (AG2F2DT) is proposed to perform smoothing and eliminate noise while preserving significant directional information. To achieve the right feature selection for optimal modeling results, the proposed Group Teaching Optimization Algorithm (GTOA) is used, so as to retain and only use significant features in modeling. The main structure of the prediction process is based on two models: the 3MDNPV3-SAN model based on multiscale dense nests, the MobileNetV3 parallel framework, and a dynamic Mantis Search Attention Network for data region emphasis. The proposed model proves itself with 99.1% accuracy, 99.3% precision, 99.7% recall, 99.6% sensitivity, 0.1% error rate, 4% computation complexity, and 0.01% computational cost while outcompeting more conventional approaches due to the incorporation of multiscale interactions as well as dynamic attention. The work hence provides a scalable robust efficient model that can be used to analyze electoral phenomena and to support decisions and design in the social sciences. The methodology is in Python which means that it is practically applicable for large databases.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.003 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.001 | 0.003 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it