The goal of this research was to explore the use of machine learning to assist in the development of knowledge-based systems (KBS) in dairy farming. A framework was first developed which described the various types of management and control activities in dairy farming and the types of information flows among these activities. This framework provided a basis for the creation of computerized information systems and helped to identify the analysis of group-average lactation curves as a promising area of application. A case-acquisition and decision-support system was developed to assist a domain specialist in generating example cases for machine learning. The specialist classified data from 33 herds enrolled with the Quebec dairy herd analysis service, resulting in 1428 lactations and 7684 tests of individual cows, classified as outlier or non-outlier, and 99 interpretations of group-average lactation curves. To enable the performance analysis of classifiers, generated with machine learning from these small data sets, a method was established involving cross-validation runs, relative operating characteristic curves, and analysis of variance. In experiments to filter lactations and tests, classification performance was significantly affected by preprocessing of examples, creation of additional attributes, choice of machine-learning algorithm, and algorithm configuration. For the filtering of individual tests, naive-Bayes classification showed significantly better performance than decision-tree induction. However, the specialist considered the decision trees as more transparent than the knowledge generated with naive Bayes. The creation of a series of three classifiers with increased sensitivity at the expense of reduced specificity per classification task, allows users of a final KBS to choose the desired tendency of classifying new cases as abnormal. For the main interpretation tasks, satisfactory performance was achieved. For the filtering tasks, performance was fai

No description available.

Seventy-seven randomly chosen herds were personally surveyed for management information. Dairymen were questioned for 1.5 hours about milking systems and practices, feeding, reproductive and genetic aspects of breeding, calf raising and finances. Seventeen non-DHI herds were included. Objectives of the study were to assemble data that described farm management practices not found in DHI records and to relate these data to four measures of profitability. Type of milking system accounted for production differences of 680 kg less milk for herds milking in flatbarns in contrast to parlors. Parlor type had no effect. Dry cow treatment reduced mastitis. A count of six recommended milking practices showed an annual increase of 246 kg milk. per cow per practice implemented. Optimal ages of first calving were found for average production, net cash income per cow and profit per cow. These were 35,29 and 26 months, respectively. Genetic indices failed to explain differences in production or profitability. Feeding programs accounted for the largest portion of expenses and were of great importance to profitability and production. Income over feed cost and feed cost as a percentage of total expense were most important, with major forage fed and degree of crop analysis having significant effects on profits. Models using investment per cow, debt per cow, average interest rate and net income per cow were good predictors of profitability (R² = .29 to .52). DHI herds were 383 kg superior to non-DHI herds in annual milk per cow, had 15 more cows and $81 more net cash income per cow, none significant. Higher production per cow led to more net cash income per cow except for medium-sized herds. / Master of Science