Predicting values in sequence

摘要

Multiple data prediction strategies are received. Each data prediction strategy may predict a next data value in a sequence of data values with a corresponding confidence value. Rather than rely on a single prediction strategy, the predictions of each of the data prediction strategies are linearly combined to generate a single prediction that is more accurate and has a lower overall loss than any of the individual prediction strategies. Further, a deviation is calculated based on the values in the sequence of values that have been observed so far using a weighted sum that favors more recent values in the sequence over less recent values in the sequence. A prediction of the next value in the sequence is generated based on the combined strategies and the calculated deviation.

主权项

1. A method comprising:
receiving a sequence of values by a computing device, wherein each value in the sequence of values has an associated time; determining a deviation of the values of the sequence of values based on the time associated with each value by the computing device; selecting a confidence function from a plurality of confidence functions by the computing device; generating a prediction based on the determined deviation and the selected confidence function by the computing device, wherein the prediction comprises a predicted next value in the sequence of values and a confidence value; providing the prediction by the computing device; and generating a bet on the next value in the sequence of values according to the prediction, wherein an amount of the bet is proportional to the confidence value.