China
Africa
Explore chapters and articles related to this topic
Test Bed Problem
Published in Vijaya Kumar Manupati, Goran D. Putnik, Maria Leonilde Rocha Varela, Smart and Sustainable Manufacturing Systems for Industry 4.0, 2023
The data was analyzed by support of three methods: (1) celeration line or trend line, (2) band with two standard deviations and (3) visual analysis. The first one was based on a reference line that expresses the general trend of the obtained data, plotted with the support of the split-middle technique, in which the baseline phase is divided into two symmetrical parts, so that the same amount of data is represented above and below the trend line. The null hypothesis assumes there is no change in the trend line between one phase and another; that is, when extending the trend line to the intervention phase, the data would also be divided equally. Statistical significance was determined by the binomial test (a = 0.05) (Portney & Watkins, 2015). The two-band standard deviation (SDB) method is based on analysis of the variability of phase data over baseline. The mean and standard deviation of this phase are calculated and, with this, two lines are drawn that delimit the range, with two standard deviations above and below the mean. These lines are extrapolated to the intervention phases. Changes between the intervention phases and baseline can be considered significant if at least two consecutive points of the intervention phase are outside the band (Portney & Watkins, 2015).
Measurement of Cognitive States in Test and Evaluation
Published in Samuel G. Charlton, Thomas G. O’Brien, Handbook of Human Factors Testing and Evaluation, 2019
The SAGAT procedure has been employed across a wide range of dynamic tasks, including aircraft control, air traffic control, driving, and nuclear power plant control (Endsley, 2000). The element common to all of them is that, by necessity, they must be tasks that can be simulated with reasonable fidelity. For each task, SAGAT questions must be developed to fully probe the situation awareness construct. Thus, probe questions should be directed at the presence or absence of elements in the situation (level 1 situation awareness), the participants comprehension of their meaning (level 2), and the anticipated future state of the elements (level 3). Question construction requires careful thought and crafting on the part of the tester. Examples of situation awareness questions presented during one SAGAT freeze in a simulated air traffic-control task are shown in Fig. 6.8 (Luther & Charlton, 1999). The number of questions presented during each freeze should be kept to a small number to minimize interference effects in working memory. The questions presented during any one freeze are thus drawn from a larger set of potential questions. Scoring participants’ answers to the questions may require establishing a tolerance band around the actual value (e.g., airspeed within 10 kn), depending on the task tolerances. As the resulting scores reflect a binomial distribution of correct/incorrect answers, a chi-square or other binomial test of significance will be required.
Selective AnDE based on attributes ranking by Maximin Conditional Mutual Information (MMCMI)
Published in Journal of Experimental & Theoretical Artificial Intelligence, 2023
Shenglei Chen, Xin Ma, Linyuan Liu, Limin Wang
To decide whether two comparing algorithms have the equal chances of win, a standard binomial sign test (Demšar, 2006) is applied to these records. Given the null hypothesis that wins and losses are equiprobable, the binomial test indicates the probability of observing the specified numbers of win and loss. In our analysis, the number of draws is divided equally to the number of wins and losses. If the resulted numbers are not integers, the number of wins will be rounded up to the next integer, while the number of losses will be rounded down to the nearest integer. We reject the hypothesis and consider the difference between the two algorithms significant if the value is less than the critical value 0.05, which is in bold font. The value we reported is the outcome of a two-tailed test. For example, value of ASAODE against AODE in terms of ZOL is 0.0011, which is the chance of observing either 49 or more wins, or 21 or fewer wins, in 70 comparisons. Since 0.0011 is less than 0.5, we can draw a conclusion that ASAODE is significantly better than AODE in terms of ZOL from the binomial sign test.