Random Contrast Learning (LuminaRCL™)
A new paradigm of machine learning.
Train models faster, with less data, on CPU.
For most applications, Lumina RCL™ outpaces neural networks by orders of magnitude in data preparation, training and inference speeds.
For every application thus far, Lumina RCL™ has outperformed state of the art deep learning neural networks.
Lumina RCL™ trains on CPU, instead of specialized hardware (GPU, TPU, NPU), and requires less compute time, power, and cooling.
Lumina RCL™ features
Simple data preparation and parameter Auto Optimization.
Unlike traditional styles of machine learning, after your data is organized into class folders, no further pre-processing or data preparation is required. Tabular data does not need to be normalized.
Our auto optimize feature applies parameters best suited for your datasets.
Less data needed.
Lumina RCL™ uses randomness as a filter to illuminate patterns as soon as they become useful and to outperform competing machine learning technologies, even when using limited training data.
Lumina’s RCLC classifier, embedded in PrismRCL™ and Lumina RCL™ API, has outperformed all competitors including Naive Bayes, SVM, Decision Trees, Neural Networks, Logistic Regression, and Random Forests in text, image, and tabular classification, while utilizing a fraction of the data, time and costs associated.
Combine models for inference.
Combine separately trained models without exposing your data in training or inference.
Random Contrast Learning (LuminaRCL™ ) is a new style of machine learning invented in 2022 by Dr. Morten Middelfart, Sam Martin, and Ben Martin, who drew upon:
- 8 years of development by Lumina;
- 40 years of pioneering work in the research and development of systems designed to detect weak signals; and
- Engagement with the philosophical tradition of phenomenology.
Train. Test. Repeat.
Time, cost of compute, and cost of specialized hardware often prohibit the iterated training of neural networks. The speed and accessibility of LuminaRCL™ via our Windows-based application and our API make experimentation more accessible.
Available in PrismRCL™ and the Lumina RCL™ API, automated parameter optimization makes rapid refinement possible.
Iterate without limitations.
PrismRCL™ Case Study: Medical Imaging
In this video, learn how you can apply PrismRCL™ to your medical imaging datasets.
At Lumina, we have classified benign and malignant tissue in breast cancer biopsy images, lymph node biopsy images, and brain cancer MRI images.