DD_Introduction

Introduction

Below is a brief description of how openHTM is organized.

Solution: OpenHTM
This is solution with all projects. It includes the core algorighm itself (Cortical Learning Algorithm), the IDE (an integrated environment to visualize the presented data in real-time, related statistics, etc) and shared projects (which include interface classes to standardize the naming of the HTM components and helper features like math functions and others).

Project: CLA (Cortical Learning Algorithm)
This is the project with the HTM core algorithm, ie the algorithm presented in the Numenta paper. It ouputs a DLL which is used by IDE project.

Project: IDE (Integrated Environment)
This is (start) project with the environment where you can choose the data to be trained, see the state of the regions/columns, statitics about previsions, and can visualize in 3d the current state of your
HTM network.

Project: IDE.BitmapProcessor
This is a sub-project of the IDE which aims to tranform the alphanumerics or geometric forms to bitmaps inputs in order to be passed to a HTM region. It ouputs a DLL which is used by IDE project.

Project: Shared.Interfaces
This is a project which provide unique interfaces to the classes used in the other projects in order to them follow a standard and so are compatible. It ouputs a DLL which is used by IDE project.

Project: Shared.Helper
This is a project which provide general functions such as math, text, etc, which are used for all projects. It ouputs a DLL which is used by IDE project.

The next pages will provide a deeper explanation about each module.

Note: Pages in construction. To understand the project, a good starting point is accessing the MainWindow.cs form in the IDE.csproj, firstly understand its function, and then "travel" the code starting by the InitializeHTM() function and StepHTMEvents() function. All code is well documented, but new comments to become it easier to understand are wellcome.