Tada

Powered by our mathematical engine, TADA builds predictive models from Small Data. TADA is to machine learning what Excel is to spreadsheet

Tada - Small Data sofware

Who are you?

You are an expert in your field, you work with data, and you must crunch numbers to make decisions.

You may not have technical skills and resources available to build predictive models.

You use spreadsheets or statistical and business intelligence tools to extract meaning from data.

You have an excellent knowledge of your product/service and market but you don’t know how to program.

Why choose TADA?

1

Easy to use

No programming and machine learning skills needed, no training required. Data preparation reduced on average by 40%

2

Easy to implement

Once built and selected, run models immediately on any platform (cloud, desktop, mobile, edge), no code conversion needed

3

Fast

Time to build and run models reduced from days/weeks down to minutes/hours

4

Competitive

Unrivalled performance on Small Data thanks to our unique mathematical expression engine inspired by evolutionary algorithms

5

Transparent

We deliver explainable & understandable models made of easy to read formulas. No black box

6

Free account

We provide you with a lifetime free account. Find out what you can accomplish with the power of MyDataModels technology.

Automated multi-platform solution

Build and run models on any platform

Features

Webapp TADA

Ease of use:

  • Web-based intuitive interface
  • Drag and drop features
  • One-click model generation
  • Data visualization and decision making tools
  • Tutorial videos at each step of the process

Data preparation:

  • Data preprocessing features: missing values, stratified sampling, time series (coming soon)
  • Automated variable reduction
  • No need to do feature engineering and manage outliers

Modeling process:

  • Creates compact predictive models made of 3 to 7 variables max
  • Optimizes speed and scalability of modeling, can be operated using a standard laptop
  • Generates a predictive model in a few minutes – complex projects that used to take months are now finished in a few hours
  • Runs in cloud and desktop environments

Deployment:

  • Deploying models for making predictions can be done with a few mouse-clicks
  • Three options are available: either via uploading your production data in Tada, or via calling a REST API endpoint (coming soon), or via exporting model’s code (Java or C++) into your application
  • Models can be run in cloud, desktop, mobile, edge and embedded environments

Modeling capabilities:

  • Binary classification modeling
  • Regression modeling
  • Multi-class classification modeling

Learn more with our resources

Get started for free

No need to download any software

Our core technology, the ZGP engine

Our technology is a unique mathematical expression engine inspired by evolutionary algorithms. We have removed some fundamental roadblocks of these algorithms enabling today their commercial use.

Understandable

Models produced are in the form of a human readable mathematical equation which can be deployed in computing languages (Java, C++). In scientific and industrial fields, having specific equations describing system behaviors enables both wide application and detailed exploration of underlying phenomena.

Webapp TADA

Minimalistic

Models produced are minimalistic in the sense of having a minimum reliance upon larger quantities of independent variables. Models have a peak maximum efficiency when employing 3 to 7 independent variables.

Discriminating

The modeling algorithm is able to consider a large number of variables during analysis and automatically select a minimal subset comprised of the most useful variables.

Efficient with Small Data

The algorithm is able to discern physical signals in small amounts of data, i.e very few rows

Facilitative

Any genotype, phenotype and evolution operator implementations intrinsically produce “search friendly” gradients when considering alternative mathematical operators, factors and independent variables. Limitations in degrees of freedom with respect to search space mobility is aggressively minimized.

Effective

Fitness function methodology that guides evolution of candidate models to generalized solutions intrinsically avoids overfit solutions. Evolutionary assessments such as tournament selection methodology is sophisticated such that fitness scores represent an estimation of performance with respect to future data not simply an evaluation of accuracy with respect to a training partition.

Insightful

The algorithm produces accurate estimation of future model performance. The accuracy realized when a model is deployed closely matches that estimated by the modeling process

Automated

The modeling process requires minimal configuration and control by a user. Modeling parameters are autonomously determined with automated analysis producing equal or superior results to those obtained by expert manual guided operation

Cross platform

Users can produce models on desktop and cloud platform. They can run models on desktop, cloud and edge devices. Supported on Windows, MacOS, and Linux

Learn more about our core technology

You work with data and you are skilled in mathematics, statistics and computer science. You know how to code (Python, R, Spark).

You prepare data, build predictive models, train them to make them as accurate as possible.

You collaborate with domain experts to support their data-driven decisions.

You perform statistical analysis, data mining and retrieval processes on data to identify trends, figures, and other relevant information. You help businesses gain a competitive edge.

You have true added value in building & testing models, not in coding. You are a problem solver.

Why choose TADA?

1

Transparent

We deliver explainable & understandable models made of easy to read formulas. No black box

2

Fast

Time to build and run models reduced to minutes/hours vs days/weeks

3

Competitive

Unrivalled performance on Small Data thanks to our unique mathematical expression engine inspired by evolutionary algorithms

4

Easy to implement

Once built and selected, run models immediately on any platform (cloud, desktop, mobile, edge), no code conversion needed

5

Easy to use

No programming or machine learning skills needed, no training required. Data preparation reduced on average by 40%.

6

Free account

We provide you with a lifetime free account. Find out what you can accomplish with the power of MyDataModels technology.

Automated multi-platform solution

Build and run models on any platform

Features

Webapp TADA

Ease of use:

  • Web-based intuitive interface
  • Drag and drop features
  • One-click model generation
  • Data visualization and decision making tools
  • Tutorial videos at each step of the process

Data preparation:

  • Data preprocessing features: missing values, stratified sampling, time series (coming soon)
  • Automated variable reduction
  • No need to do feature engineering and manage outliers

Modeling process:

  • Creates compact predictive models made of 3 to 7 variables max
  • Optimizes speed and scalability of modeling, can be operated using a standard laptop
  • Generates a predictive model in a few minutes – complex projects that used to take months are now finished in a few hours
  • Runs in cloud and desktop environments

Deployment:

  • Deploying models for making predictions can be done with a few mouse-clicks
  • Three options are available: either via uploading your production data in TADA, or via calling a REST API endpoint (coming soon), or via exporting model’s code (Java or C++) into your application
  • Models can be run in a cloud, desktop, mobile, edge, and embedded environments

Modeling capabilities:

  • Binary classification modeling
  • Regression modeling
  • Multi-class classification modeling

Learn more with our resources

Get started for free

No need to download any software

Our core technology, the ZGP engine

Our technology is a unique mathematical expression engine inspired by evolutionary algorithms. We have removed some fundamental roadblocks of these algorithms enabling today their commercial use.

Understandable

Models produced are in the form of a human-readable mathematical equation which can be deployed in computing languages (Java, C++). In scientific and industrial fields, having specific equations describing system behaviors enables both wide application and detailed exploration of underlying phenomena.

Webapp TADA

Minimalistic

Models produced are minimalistic in the sense of having a minimum reliance upon larger quantities of independent variables. Models have a peak maximum efficiency when employing 3 to 7 independent variables.

Discriminating

The modeling algorithm is able to consider a large number of variables during analysis and automatically select a minimal subset comprised of the most useful variables.

Efficient with Small Data

The algorithm is able to discern physical signals in small amounts of data, i.e very few rows

Facilitative

Any genotype, phenotype and evolution operator implementations intrinsically produce “search friendly” gradients when considering alternative mathematical operators, factors and independent variables. Limitations in degrees of freedom with respect to search space mobility is aggressively minimized.

Effective

Fitness function methodology that guides the evolution of candidate models to generalized solutions intrinsically avoids overfit solutions. Evolutionary assessments such as tournament selection methodology is sophisticated such that fitness scores represent an estimation of performance with respect to future data not simply an evaluation of accuracy with respect to a training partition.

Insightful

The algorithm produces an accurate estimation of future model performance. The accuracy realized when a model is deployed closely matches that estimated by the modeling process

Automated

The modeling process requires minimal configuration and control by a user. Modeling parameters are autonomously determined with automated analysis producing equal or superior results to those obtained by expert manual-guided operation

Cross-platform

Users can produce models on a desktop and cloud platform. They can run models on desktop, cloud and edge devices. Supported on Windows, MacOS, and Linux

Learn more about our core technology

TADA Packages

Choose the package that best suits your needs

Free Edition

Free lifetime
account

€0

  • Cloud
  • Single user
  • Project creation – Dataset CSV support until 10 Mb
  • Model generation – 20 models
  • Score generation – 20 scores
  • Use cases
  • FAQ support

Premium Edition

1 year
license

€ 2 950

  • Cloud
  • Single user
  • Project creation – Dataset CSV support until 30 Mb
  • Model generation – 50 models per day
  • Score generation – 50 scores per day
  • Use cases & Dataset samples
  • Pre-processing features
  • Source code to execute models
  • FAQ & email support

Pro Edition

1 year
license

From € 12 500

  • Cloud & desktop
  • Single user
  • Project creation – Dataset CSV support until 50 Mb
  • Model generation – Unlimited
  • Score generation – Unlimited
  • Feature reduction
  • Use cases & Dataset samples
  • Pre-processing features
  • Source code to execute models
  • FAQ & custom support
  • Packaged Offers and Packaged PoC Offers
  • Training

Enterprise Edition

Unlimited
enterprise access

On demand

  • Private / Public Cloud & desktop
  • Multi-user
  • Sharing mode
  • Unlimited file system support
  • Unlimited models & scores
  • Feature reduction
  • Use cases & Dataset samples
  • Pre-processing features
  • Source code to execute models
  • Technical support via email with SLA
  • Consulting & custom support
  • Training