Machine Learning Cores

Machine learning refers to a system capable of the autonomous acquisition and integration of knowledge. This capacity to learn from experience, analytical observation, and other means, results in a system that can improve its own speed or performance over time.

A major focus of machine learning research is to automatically learn to recognize complex patterns and make intelligent decisions based on data. The difficulty lies in the fact that the set of all possible behaviors given all possible inputs is too large to be covered by the set of observed examples (training data). Hence the learner must generalize from the given examples, so as to be able to produce a useful output in new cases.

One measure of progress in machine learning is its significant real-world applications, such as: machine perception, computer vision, object recognition, search engines, medical diagnosis, bioinformatics, brain-machine interfaces, stock market analysis, speech and handwriting recognition, robot control and robot locomotion, etc.

There are many different predictive models (classifiers) in machine learning, the most popular being Artificial Neural Networks (ANNs), Decision Trees (DTs) and Support Vector Machines (SVMs). Recently, a new way of making more accurate predictive models has emerged, called ensemble learning.

General characteristic of all machine learning approaches is that they are computationally intensive requiring powerful computers to execute them. This fact makes the usage of machine learning systems in embedded devices prohibitive. On the other hand, upcoming vision of ambient intelligence, where electronic devices work in concert to seamlessly support people in carrying out their everyday life activities, tasks and rituals in easy, natural way, will require embedding some sort of intelligence within embedded systems. This will only be possible if machine learning systems are implemented directly in hardware, as IP cores.

So-Logic is one of the first companies to have addressed this need and developed a range of IP cores that enable integration of machine learning systems into FPGA-based embedded systems today.

Our portfolio of machine learning IP cores includes:

• Decision Trees Cores - IP cores that can be used to implement previously designed decision tree directly in hardware, or even infer decision tree using supplied training data
• Ensemble Classifiers Cores - IP cores that enable the hardware implementation of ensemble classifier systems. Here you can find IP cores that can be used to:
  • Ensemble Evaluation Cores - IP cores that can be used to calculate the predictions of individual members of the ensemble
  • Combination Rules Cores - IP cores that are used to combine the predictions of individual members in order to make a collective decision
  • Ensemble Inference Cores - IP cores that are able to create ensemble classifiers from supplied training data