- [Changed] Project renamed : Korea University Conditioning Simulator
- [Changed] model notation is improved. (does not effect the script)
- Program's equation image is updated
- [Changed] Graph drawing speed is significantly increased in some devices (drawing time reduced 80%)
- I don't know why, but legend function was consuming a lot of time. (preallocation problem?)
- [Changed] SPH model doesn't update alpha if the corresponding CS is not presented.
- The original thesis does not clearly specify this in the equation.
- However, if this rule is not applied to the alpha updating rule, all alpha values remain the same across CSs.
- [Fixed] Pearce-Hall model error fixed (I'm keep writing it as "Pearson-Hall")
- [Changed] Default model parameter changed
- [Fixed] weight is now initialized everytime the simulation is performed.
- [Changed] Since the initial alpha values are presented in the 'Stage Configuration' panel, now, it's not presented in the 'Experiment Schedule' panel.
- [Changed] Experiment Schedule notation changed.
- Deleted alpha values
- No US is now encoded as '-'
- Experiment schedule parsing method is also changed.
- [Changed] Mackintosh model is changed into the new form.
- applyed (1-a) and (a) term in the delta alpha. This smooths the increase/decrase curve and automatically limits the alpha in [0,1].
- New equation is similar to the formalization of Moore & Stickney(1975).
- But it uses small negative number 'epsilon' when the discrepancy of the CS and X are equal.
- This change allows the model to successfully predict Latent Inhibition.
- [Changed] All Experiment Template were changed to have initial alpha value of 0.5. (Previously, there were two options).
Initial working version