NumSharp Cheat Sheet

NumPy to NumSharp Cheat Sheet

Recently, I published a small example project to utilize the htm.core AI algorithm by consuming its REST API via C#. As the API also transfers serialized multi-dimensional NumPy arrays, I was looking for an easy way to get them back into C# objects. I’ve tried out a couple of approaches and finally decided on using the NumSharp library, as I wanted a solution that works on multiple platforms.

I find DataCamps’ Data Science Cheat Sheets very useful and was hoping to find something similar for NumSharp. Well, I didn’t, but obviously that gave me a … more

htm.core parameters – Single Order Sequence Memory

To allow the htm.core temporal memory to learn sequences effectively, it is import to understand the impact of the different parameters in more detail.

In this part I will introduce

  • columnDimensions
  • cellsPerColumn
  • maxSegmentsPerCell
  • maxSynapsesPerSegment
  • initialPermanence
  • connectedPermanence
  • permanenceIncrement
  • predictedSegmentDecrement

Temporal Memory – Previously on this blog…

Part 1 just covered enough basics of htm.core to get us started, and we actually saw how the single order memory got trained.
A cycle of encoded increasing numbers from 0 to 9 was very easy to predict, as there was always just one specific value that could follow the … more

Using the htm.core-jupyter docker image


To make it easier to get started with some of my htm.core experiments or with htm.core in general, I thought it would make sense to provide a docker image with htm.core preinstalled. So here it is:
Please welcome the htm.core-jupyter image.

This image is using the scipy-notebook as foundation, with the htm.core package installed on top of it.

For a list of the other preinstalled python packages, just look here.
There are also a lot of other Jupyter docker images available. I recommend starting with the Jupyter Docker Stacks quick start page for a … more

Hierarchical Temporal Memory – part 1 – getting started

Hierarchical Temporal Memory - Neuron


I came across the concept of Hierarchical Temporal Memory (HTM) and its implementation a while ago, and am still very fascinated about this approach to artificial intelligence.

When, about one year ago, the active development shifted towards the community fork named htm.core, which supports Python 3, it became finally time to have a closer look and try it out by myself.

BTW: According to this forum post, there are no plans to upgrade the older NuPIC library to Python 3.

A lot of documentation about the theory of HTM is available at numenta.orgmore

Apache Zeppelin mit PySpark und PostgreSQL benutzen – Teil 1

In dieser kleinen Artikelserie beschreibe ich, wie man Apache Zeppelin mit dem PySpark Interpreter benutzen kann, um eine PostgreSQL Datenbank abzufragen.

Im ersten Teil werde ich das offizielle Docker Image von Apache Zeppelin 0.8.1 um Spark 2.4.0 erweitern. Danach geht es dann weiter mit der Konfiguration des Spark Interpreters um auf PostgreSQL zugreifen zu können. Anschließend werden wir uns mit PgBench eine Datenbank mit Test Daten generieren und zu guter Letzt werden wir diese Daten mit PySpark abfragen

Beginnen wir also mit dem Erweitern des Docker Images. Dazu habe ich folgendes Dockerfile verwendet:

FROM apache/zeppelin:0.8.1

Scroll to top