Parameter estimation of the hyper-geometric distribution model for real test/debug data

Framework for modeling software reliability, using various testing-efforts and fault-detection rates

“…1) Constant TE Consumption: In the derivation of classical SRGM [4]- [6], [11]- [14], [17], [22], most researchers assumed that the TE (workload) of a software system is constant: 42) Weibull-Type TE Function: According to [1]- [3], [5], [23], TE should not be assumed constant throughout the testing phase. In fact, instantaneous TE ultimately decreases during the testing life cycle so that the cumulative TE approaches a finite limit.…”

“…1) Constant Proportionality: Most existing SRGM assume that the mean number of faults detected in is proportional to the number of remaining faults [4], [11]- [14], [17], [22]: 12is a constant proportionality.…”

The discrete time hyperexponential model for software reliability growth evaluation

“…They will be referred to as data set A [31] and data set B [28] respectively. The model will be applied in order to follow up the evolution of cumulative number of failures.…”

“…The failure data reported in [28] and shown on Fig. 7 were recorded during combinational test of a real field software: 418 executions were performed during which 137 failures were observed.…”

A fast algorithm to compute maximum likelihood estimates for the hypergeometric software reliability model

“…This model has been studied in a number of papers [5][6][7][8][9][10][11][12][13][14][18][19][20][21]. In the hypergeonietric model, the result of software test k consists of the number I U~ of faults detected in the test and the number zk of those faults which are newly detected.…”