26th International Conference on Inductive Logic Programming

4th - 6th September 2016, London


2016

L

P






Competition

Probabilistic track: results

There was a single entry in the probabilistic track of the competition. The entry used the SLIPCOVER system and was by Riccardo Zese, Elena Bellodi and Fabrizio Riguzzi of the University of Ferrara. For details of this entry see this document.

A break down of the results is shown below. To see how the system performed on various types of problem, you can use the checkboxes. The table, and average penalty will update accordingly.

Note: In the probabilistic track, the penalty paid is the squared difference between the true probability of a trace, and the probability returned by the system.

Include Problems without:         
Include Problems with:         
     
Tasks with small language biases, are tasks where the list of relevant predicates given to the learner contain only the predicates needed to explain the examples. Medium sized language biases contain an extra, irrelevant predicate, and large language biases contain at least two extra irrelevant predicates. Note that in tasks that required predicate invention, the invented predicates were not given to the learner.

Note that in the table below, the list of properties is exhaustive. So, for example, if a problem does not have recursion in its list of properties, then it is possible to explain the examples without recursion. You can see details of the problem, together with the test traces and the final attempt made by the learner to estimate the probabilities of these traces by clicking on the "problem details and system output" links.

Problem Name Properties Average penalty Details
Self destruct probabilistic; predicate invention; small language bias 0.10731910869117492 problem details and system output
Self destruct probabilistic; predicate invention; medium language bias 1.0 problem details and system output
Self destruct probabilistic; predicate invention; large language bias 0.03766469571413892 problem details and system output
security camera probabilistic; non observational predicate learning; non monotonic; small language bias 1.0 problem details and system output
security camera probabilistic; non observational predicate learning; non monotonic; medium language bias 1.0 problem details and system output
security camera probabilistic; non observational predicate learning; non monotonic; large language bias 1.0 problem details and system output
connections probabilistic; small language bias 0.43052954036740737 problem details and system output
connections probabilistic; medium language bias 0.43059400508573586 problem details and system output
connections probabilistic; large language bias 0.02799999999999996 problem details and system output
probabilistic lock cell probabilistic; non observational predicate learning; non monotonic; small language bias 0.5246192511301718 problem details and system output
probabilistic lock cell probabilistic; non observational predicate learning; non monotonic; medium language bias 0.5343646107651215 problem details and system output
probabilistic lock cell probabilistic; non observational predicate learning; non monotonic; large language bias 0.5114627140165431 problem details and system output
stun probabilistic; predicate invention; non monotonic; recursion; small language bias 0.16649327715311013 problem details and system output
stun probabilistic; predicate invention; non monotonic; recursion; medium language bias 0.20608557101112718 problem details and system output
stun probabilistic; predicate invention; non monotonic; recursion; large language bias 0.421630859375 problem details and system output
nonOPL stun probabilistic; non observational predicate learning; non monotonic; recursion; small language bias 0.220458984375 problem details and system output
nonOPL stun probabilistic; non observational predicate learning; non monotonic; recursion; medium language bias 0.220458984375 problem details and system output
nonOPL stun probabilistic; non observational predicate learning; non monotonic; recursion; large language bias 0.220458984375 problem details and system output
boost probabilistic; non observational predicate learning; recursion; small language bias 0.35963541666666665 problem details and system output
boost probabilistic; non observational predicate learning; recursion; medium language bias 0.35963541666666665 problem details and system output
boost probabilistic; non observational predicate learning; recursion; large language bias 0.35963541666666665 problem details and system output
boost: predicate invention probabilistic; predicate invention; recursion; small language bias 0.3101159160482062 problem details and system output
boost: predicate invention probabilistic; predicate invention; recursion; medium language bias 0.1789400175183276 problem details and system output
boost: predicate invention probabilistic; predicate invention; recursion; large language bias 0.319580078125 problem details and system output
tripwire probabilistic; non observational predicate learning; predicate invention; non monotonic; small language bias 1.0 problem details and system output
tripwire probabilistic; non observational predicate learning; predicate invention; non monotonic; medium language bias 1.0 problem details and system output
tripwire probabilistic; non observational predicate learning; predicate invention; non monotonic; large language bias 1.0 problem details and system output
tripwire with off switch probabilistic; non observational predicate learning; predicate invention; non monotonic; recursion; small language bias 1.0 problem details and system output
tripwire with off switch probabilistic; non observational predicate learning; predicate invention; non monotonic; recursion; medium language bias 1.0 problem details and system output
tripwire with off switch probabilistic; non observational predicate learning; predicate invention; non monotonic; recursion; large language bias 1.0 problem details and system output
recursive valid move with walls probabilistic; non monotonic; small language bias 0.06187884577600208 problem details and system output
recursive valid move with walls probabilistic; non monotonic; medium language bias 0.0618794026576696 problem details and system output
recursive valid move with walls probabilistic; non monotonic; large language bias 0.06187808329255562 problem details and system output
invalid_move probabilistic; non observational predicate learning; non monotonic; small language bias 0.26881182949292803 problem details and system output
invalid_move probabilistic; non observational predicate learning; non monotonic; medium language bias 0.24485829943399126 problem details and system output
invalid_move probabilistic; non observational predicate learning; non monotonic; large language bias 0.3321287980344474 problem details and system output
recursive valid move: nonOPL probabilistic; non observational predicate learning; recursion; small language bias 0.12799999999999984 problem details and system output
recursive valid move: nonOPL probabilistic; non observational predicate learning; recursion; medium language bias 1.0 problem details and system output
recursive valid move: nonOPL probabilistic; non observational predicate learning; recursion; large language bias 0.07890820312499987 problem details and system output
energy probabilistic; non observational predicate learning; small language bias 1.0 problem details and system output
energy probabilistic; non observational predicate learning; medium language bias 1.0 problem details and system output
energy probabilistic; non observational predicate learning; large language bias 1.0 problem details and system output
locking cells probabilistic; predicate invention; non monotonic; recursion; small language bias 0.0241142473627678 problem details and system output
locking cells probabilistic; predicate invention; non monotonic; recursion; medium language bias 0.0241142473627678 problem details and system output
locking cells probabilistic; predicate invention; non monotonic; recursion; large language bias 0.14969771934235576 problem details and system output