Brainome Data Compiler
Brainome is the world’s first data compiler for solving supervised machine learning problems.
Benefits
✓ Training is 100 times faster.
✓ Models are 1000 times smaller.
✓ Handle data sets of any size.
✓ Run locally and keep your data private.
✓ Self contained models that are 100% portable.
Customers choose Brainome when they:
Brainome is a data compiler
Just like Software Compilers translate source code into binary code...
Source Code
GNU C Compiler
Binary Code
... a Data Compiler translates data into predictive models.
Data Set
Data Compiler
Model
Brainome treats data as its “source code”
Features and Requirements:
Input Data Set
- CSV Format
- One column must contain the class labels (target column)
- Cell values supported: strings, floats, integers
- No pre-processing necessary
- No limit on number of rows or columns
- Unlimited number of classes. 100 instances per class recommended for best results.
- Support for unbalanced data sets
- Support for sparse data sets
- Support for missing value data sets
Model Creation
- Select from Random Forest, Neural Network, Support Vector Machine and Decision Tree models
- Measurement driven building process for optimum model size and speed
- Produces very small models, often kilobytes in size; 2 to 3 orders of magnitude smaller than models produced by other tools
- Stand-alone python executable that requires only Numpy
- Written in clear text Python code (easily committed to your Git repository) GPU not required to run model
Measurements
- Model Generalization
- Memory Equivalent Capacity
- Attribute Ranking
- Risk of Overfit
- Data Sufficiency
Integrations
- Brainome can be scripted from the command line
- Brainome measurements are encapsulated in JSON or plain text
