KDnuggets Home » News » 2015 » Mar » Software » Automatic Statistician is here: Dr. Mo ( 15:n09 )

Automatic Statistician is here: Dr. Mo

Dr. Mo, Automatic Statistician is here! Using Artificial Intelligence, self-learning algorithm, multimodel technology Dr. Mo achieves Super Accuracy and Speed. Simple use and simple output for non-statisticians.

By Dr. Milorad Krneta (soft10ware).

I had my first encounter with Big Data when working on a UN Population Fund project called “Bachue” more than 30 years ago.

To finish the assignment, I had to automate linear equations, which I did that by using Artificial Intelligence (AI) basic theory. In that process, I realized that it I had a self-learning algorithm which would cover not only numerical variables but any type of variables.

Pretty soon, it became evident that the parametric statistics was not a viable option for non-numeric variables. Therefore, the solution was in the probability space of non-parametric statistics.

In the academia and science world, I developed the theory of self-learning in order develop automatic statistical software: which he applied with success.

The first self-learning practical application was in clustering and decision tree technology which I developed in mid 90’s. The software was able to use any type of variables.

The next software I developed, was an Automatic Decision-Tree technology which was using multiple interaction detection technology instead of a binary-split approach.

Understanding that “real world” would be the most exciting place to apply self-learning, Artificial Intelligence statistical software, I moved from university to the private enterprise world.

In my first company, Generation5, I developed the first commercial, Automatic Statistical Software called MWM. MWM was proved to more than 80 Blue Chip Companies to be most the effective prediction “Brain”.

dr-mo-automatic-statistician In my second company “Infer systems”, I developed an Automatic Statistician for Real-Time-Bidding. Both companies were bought out and are successfully still doing business today.

In my third company, I developed a new Automatic Statistician, Dr. Mo, which is not only fast but also much more accurate than any human or manual method on the market today.

This success was achieved by introducing Automatic Multi-Modelling technology. The predictions are no longer constrained by the Best, “One model” technology: and this has brought a never seen before increase in prediction accuracy.

In addition, by creating Artificial Intelligence which does not require any transformation of original variables- it has eliminated the variable transformation cause of prediction error.

I was always guided by the principles of efficiency, simplicity, accuracy and reliability principles.

I developed Dr. Mo – Automatic Statistician to be:
  • Automatic: Artificial Intelligence, Self-Learning Statistical Software
  • Super-Fast: Able to use Any database Size or Big Data
  • Super Accurate: Multi-Model, Non-Parametric Technology
  • Data Agnostic: Use any type of variables
  • No Over fitting: Keeping models simple
  • User Friendly: Non-Statistician can use it immediately
  • Actionable: Prediction output simple to understand
  • Guaranteed: No other more accurate or faster predictions
  • Simple-to-use: Does not require statistical knowledge to understand it
  • Validation: Non-statisticians can validate any model produced
  • Reliability: Never produces insignificant predictions

Beside common types of one variable predictions, Dr. Mo has been used for multiple variables predictions and, even more important, for longitudinal predictions.

Dr. Mo has been successfully been tested on more than 267 different databases with applications ranging from: finding the best customer to predicting gene sequencing for particular proteins to predicting healthcare needs by region, to consumer or political preferences etc..

With the ability to use Big data with great accuracy and speed like never before, I hope that Dr. Mo will be applied to help prevent or cure diseases, to preventing world poverty to help businesses thrive by making more accurate and best decisions!

Bio: Dr. Milorad Krneta Dr. Milorad Krneta started his science and university career in former Yugoslavia at Belgrade University. In his latest firm, Soft10 Inc., Dr. Krneta developed automatic prediction software, called Dr. Mo, Automatic Statistician.


Sign Up

By subscribing you accept KDnuggets Privacy Policy