3 papers
cs.LG2018
Unreasonable Effectivness of Deep Learning
Finn Macleod
We show how well known rules of back propagation arise from a weighted combination of finite automata. By redefining a finite automata as a predictor we combine the set of all -…
math.ST2009
Universal Relationships in Measures of Unpredictability
Finn Macleod, Alexei Pokrovskii, Dima Rachinskii
The predictability of a sequence is defined as the asymptotic performance of the best performing predictor in a given class. The value of the predictability of a sequence will in g…
stat.ML2008
Prediction with Restricted Resources and Finite Automata
Finn Macleod, James Gleeson
We obtain an index of the complexity of a random sequence by allowing the role of the measure in classical probability theory to be played by a function we call the generating mech…