On the efficiency of OLS reduced probabilistic neural networks for aircraft-flare discrimination
Bibliographic record
Abstract
Probabilistic neural networks (PNN) are the instruments of choice when it comes to critical decision making. Indeed, their output is not a simple yes-or-no decision; they are able to produce the probability that the features received as their input correspond to an object of any one of many classes. In work reported elsewhere, we have devised such a network for the discrimination of aircrafts from their decoy flares. It is very efficient, consistently exhibiting a recognition success rate of the order of 98-99%. However, because these neural networks are based on the Parzen-windows method, they must contain a very large number of neurons in order to be efficient. This can represent a serious disadvantage when they are to be incorporated in a real time system. It is thus advantageous to be able to reduce their size, without affecting appreciably their performance. We report in this article on the success we have had with adapting and applying an Orthogonal Least Squares (OLS) reduction method to the probabilistic neural network we built previously. We show that this method allows for a reduction of the number of neurons by as much as 81.9% with a decrease in performance of only 0.6%. Even a drastic reduction of 97.7% of number of neurons still produces a network with a 93.5% success rate. A side benefit of the application of this method to PNNs, is an ordered list of the images that the neural network considers as the best representatives of their class of objects.
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.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.008 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 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 itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".