# Applications **URL:** https://research.allora.network/c/applications/5.md [Latest](https://research.allora.network/latest.md) · [Categories](https://research.allora.network/categories.md) · [Tags](https://research.allora.network/tags.md) --- ## [About the Applications category](https://research.allora.network/t/about-the-applications-category/23)
**Author:** [@Apollo11](https://research.allora.network/u/Apollo11)\ **Replies:** 0\ **Last updated:** [May 28, 2024, 4:33pm UTC](https://research.allora.network/t/about-the-applications-category/23 "2024-05-28T16:33:11Z")
Research aimed at developing specific applications of Allora. This category should be used for general AI/ML model development, identifying and developing suitable use cases for the network, and identifying areas where … --- ## [CZAR loss function for returns prediction topics](https://research.allora.network/t/czar-loss-function-for-returns-prediction-topics/155)
**Author:** [@joel](https://research.allora.network/u/joel)\ **Replies:** 4\ **Last updated:** [February 9, 2026, 2:38am UTC](https://research.allora.network/t/czar-loss-function-for-returns-prediction-topics/155 "2026-02-09T02:38:51Z")
The CZAR (Composite Zero-Agnostic Return) loss function is designed to address limitations observed in previous loss formulations used for inference synthesis in returns prediction topics. This post documents the motivat… --- ## [Losses in returns prediction topics](https://research.allora.network/t/losses-in-returns-prediction-topics/120)
**Author:** [@joel](https://research.allora.network/u/joel)\ **Replies:** 8\ **Last updated:** [November 4, 2025, 10:12am UTC](https://research.allora.network/t/losses-in-returns-prediction-topics/120 "2025-11-04T10:12:43Z")
I’ve noticed some strange behaviour of the losses in topics predicting log-returns. Some workers occasionally provide extremely large inferences (up to ~ 10^12) compared to typical returns values (\<0.1), but still have r… --- ## [Feature engineering experiments 1: add log-returns-focused features](https://research.allora.network/t/feature-engineering-experiments-1-add-log-returns-focused-features/134)
**Author:** [@Apollo11](https://research.allora.network/u/Apollo11)\ **Replies:** 8\ **Last updated:** [October 1, 2025, 5:59pm UTC](https://research.allora.network/t/feature-engineering-experiments-1-add-log-returns-focused-features/134 "2025-10-01T17:59:09Z")
In this experiment, we test the impact of adding a returns-focused feature set (all quantities you can calculate for price, but for log-returns). This is a relatively small amount of work (applying the transformations yo… --- ## [Price/returns topic feature engineering](https://research.allora.network/t/price-returns-topic-feature-engineering/123)
**Author:** [@Apollo11](https://research.allora.network/u/Apollo11)\ **Replies:** 14\ **Last updated:** [July 17, 2025, 12:18pm UTC](https://research.allora.network/t/price-returns-topic-feature-engineering/123 "2025-07-17T12:18:08Z")
Looking at participant performance in Allora’s Forge programme, I have the impression that feature engineering is the bottleneck in several of the current price/return prediction topics. Markets are like physical system… --- ## [Thorough testing of new forecaster model](https://research.allora.network/t/thorough-testing-of-new-forecaster-model/117)
**Author:** [@joel](https://research.allora.network/u/joel)\ **Replies:** 9\ **Last updated:** [May 29, 2025, 11:55am UTC](https://research.allora.network/t/thorough-testing-of-new-forecaster-model/117 "2025-05-29T11:55:49Z")
Forecasting is the component that enables context-awareness in the Allora Network. During inference synthesis, worker inferences are combined using their historical performance (through an exponential moving average of t… --- ## [Hybrid ML models](https://research.allora.network/t/hybrid-ml-models/103)
**Author:** [@Ryouhei-Deworkhub](https://research.allora.network/u/Ryouhei-Deworkhub)\ **Replies:** 1\ **Last updated:** [January 13, 2025, 2:02pm UTC](https://research.allora.network/t/hybrid-ml-models/103 "2025-01-13T14:02:15Z")
We aim to identify an effective approach to constructing robust predictive models. It is well-recognized in the academic community that, in addition to single models, hybrid models have also garnered significant attentio… --- ## [Standard forecaster model](https://research.allora.network/t/standard-forecaster-model/37)
**Author:** [@Apollo11](https://research.allora.network/u/Apollo11)\ **Replies:** 9\ **Last updated:** [November 27, 2024, 4:40pm UTC](https://research.allora.network/t/standard-forecaster-model/37 "2024-11-27T16:40:22Z")
We will need to develop an “off-the-shelf” forecaster model that uses a combination of: network data (e.g. historical inference losses, raw inferences, network inferences, worker scores, worker rewards); private data (…