student work |
review notes

research materials |
tutorials | |

home to: course | instructor |
department | La Salle home

E-Mail: redmond@lasalle.edu  

Some notes on CLIPS: certainly a draft at this point

Some CLIPS examples: zoo animals rules (right click to download) Use to infer animal in one of the following: first animal or second animal or third animal or fourth animal (right click to download)

media advisor rules (right click to download)

relatives rules for CLIPS (right click to download) and relatives basic startup info rule for CLIPS (right click to download)

NJ crime rules (right click to download) Used to illustrate templates in CLIPS. Use to infer crime level in one of the following: first city or second city or third city

Zebra puzzle rules (right click to download) Uses templates in CLIPS (but is also pretty hairy).

london weather prediction rules for FuzzyCLIPS (right click to download) and london weather prediction basic startup info rule (one problem) for FuzzyCLIPS (right click to download)

FuzzyCLIPS executable (right click to download) (to be used only if the system hasn't been installed)

Very Simple Speed Control Simulation Fuzzy Prediction rules (right click to download) Used to illustrate Fuzzy Logic in FuzzyCLIPS. Trace from one problem: too fast (problem built into the above file)

Spare Parts Business rules (right click to download) Used to illustrate Fuzzy Logic in FuzzyCLIPS. Use to recommend number of spare parts to start with for one of the following: first set of assumptions or second set of assumptions (THIS IS THE BETTER EXAMPLE). Traces gathered from these: first set of assumptions or second set of assumptions

London Weather Fuzzy Prediction rules (right click to download) Used to illustrate Fuzzy Logic in FuzzyCLIPS. Use to predict tomorrows rainfall for one of the following: first day or second day Traces gathered from these: first day or second day

Shower Simulation Fuzzy Templates (right click to download) (when loading, load first) Shower Simulation Functions (right click to download) (when loading, load second) Shower Simulation Fuzzy Prediction rules (right click to download) (when loading, load third). Used to illustrate Fuzzy Logic in FuzzyCLIPS. Fairly complicated. It requests input from user, so many different problems can be tried. Trace from one: starts out too cold and too strong

I do not have a sample test from this class. But to give an idea of the kind of tests I like to give, here is one from a data mining class not too long ago: data mining midterm 2004 WE HAVEN'T COVERED ANY OF THIS MATERIAL!!! SO DON'T LOOK AT THIS FOR content, LOOK AT IT FOR style!!!!

Supplemental PowerPoints for Case-Based Reasoning

Some notes on using the AIAI CBR Tool: certainly a draft at this point

Some files for playing with AIAI CBR Tool: (Internet Explorer doesn't know anything about these, so right click to download)

Zoo - animals again data description (key) file (load first) training data (cbr) file (load second) test data (tst) file (load third)

NJ Crime again data description (key) file (load first) training data (cbr) file (load second) test data (tst) file (load third)

Basketball - predicting high scorers (pts per minute - nominalized - using numeric inputs) data description (key) file (load first) training data (cbr) file (load second) test data (tst) file (load third)

Supposedly from a bank in Japan (approve loan or not) - getting into a typical data mining task (if you had a lot more data) data description (key) file (load first) training data (cbr) file (load second) test data (tst) file (load third)

Some Excel spreadsheets illustrating neural network techniques on simple examples (right click to download)

Perceptron learning of AND (initial weights as in book example) Back Propagation learning of Xor (initial weights as in book example) (note slow progress)

Some files for playing with Weka Machine Learning Tool: (Internet Explorer doesn't know anything about these, so right click to download)

Some simple logical examples: AND (easy) and Xor (not so easy) and Xor (not so easy) and Xor (many repetitions makes easier to learn )

Some old friends: NJ crime nominal

basketball numerics with class to be predicted

japan bank

broadway shows

For Genetic Algorithms, a trace of a simple genetic algorithm solving the power generation maintenance scheduling problem posed in the Negnevitsky book.

Zip file containing code and files for a simple genetic algorithm to solve the power generation maintenance scheduling problem posed in the Negnevitsky book.

Instructions for using the software above (requires Java and an IDE)

Some Excel spreadsheets related to genetic algorithm topics (right click to download)

Complex mathematical function - encoding and evaluation (calculation checking book example) Power unit example - checkig to see if genetic algorithm found optimal solution (book's 3rd example)

Decision Trees Decision Tree Gain (showing difference between machine generated splits and human chosen splits - book example)