TRACX simulator

Truncated Recursive Autoassociative Chunk eXtractor

TRACX is a new model of sequence learning. It is a connectionist autoassociator model which fits a wide range of phenomena from the infant statistical learning and adult implicit learning literature. TRACX outperforms PARSER (Perruchet & Vintner, 1998) and the simple recurrent network (SRN, Cleeremans & McClelland, 1991) in matching human sequence segmentation on existing data. This simulator is written entirely in Javascript and runs locally in your browser. A short summary of the algorithm can be found here and more details of the model can be found in:

French, R. M., Addyman, C., & Mareschal, D. (2011). TRACX: A recognition-based connectionist framework for sequence segmentation and chunk extraction Psychological Review, 118(4), 614–636. doi:10.1037/a0025255 [pdf - 872k]

Step 1 - Choose Training data:

Sentences should each begin on a new line or be separated by commas.
A short explanation of the original study will appear here.
Step 2: Set Network Parameters
Learning Rate
Recognition Criterion
Reinforcement Probability
Momentum
Sentence repetitions
Temperature
Random Seed
Input encoding Local
Distributed (binary)
Delta rule Absolute maximum error
Root mean square error
Step 3: Run Simulator
Training Step 0
Batch Mode Number of subjects:
The internal parameters used by the network will appear here.
Network Activations
Network Error
Words: Part Words: Non Words:
Test words.. ..every N steps:On/Off

Test items should be separated by commas.