PROF. BADAMANA MOHAMED SALIM (RETIRED)

A comparison of productivity and adaptability of indigenous (Boran and Small East African Zebu) and the exotic (Sahiwal and Ayrshire) cattle on Kenyan ranches located in semi-arid areas of the Rift Valley Provinces was done. Data sets of the cattle breeds over the 1979-1993 period on Deloraine, Elkarama, Ilkerin, National Sahiwal Stud (NSS) and Oljorai were analyzed by the least squares fixed effects model procedures. The least squares means of productivity component traits of each breed were imputed in PRY model to derive feed energy efficiency (FEE) index and carry out sensitivity analyses. The respective FEE indices were 137, 106, 100, 86, 82, 79, 78, and 65 aggregate value off take per unit of feed energy requirements for NSS Sahiwal, Deloraine Ayrshire, Deloraine Sahiwal, Elkarama Sahiwal, Ilkerin Sahiwal, Elkarama Boran, Oljorai Boran and Ilkerin small East African Zebu. The sensitivity showed that survival and reproductive (fitness) traits of the indigenous cattle were lower than for exotic cattle indicating better adaptability of indigenous cattle. The sensitivity values of milk yield, mature weight and mature age were at near optimal levels for the ranch environments, except for Deloraine Ayrshire and Elkarama Boran. The results suggest that fitness traits and rate of genetic progress by selection in production traits could be further enhanced by improving environmental management on the ranches. The implications on development strategies of indigenous and exotic cattle are also discussed.

Knowledge discovery in databases (KDD) should provide not only accurate predictions but also comprehensible rules. In this paper, we demonstrate that the machine learning approach of rule extraction from a computer trained neural network system can successfully be applied to milk production analyses in dairy cattle. Such extracted knowledge should be useful in interpretation and understanding how the neural network (NN) model makes its decision. Data consisting of 6095 lactation records made by cows from 76 officially milk recorded Holstein Friesian herds in the period 1988-2005 were used to extract rules using neural network. Two different methods of attribute categorization; auto-class and the domain expert were used. For automated knowledge acquisition, rule induction used Weka software while SAS was used in domain expert. The neural nets were first trained to identify outputs for different inputs. The trained networks were then used for rule extraction. The study showed that the decision trees generated from the trained network had higher accuracy than decision trees created directly from the data. The study also indicated a need for a process to determine important inputs before using a neural net and showed that reduced input sets may produce more accurate neural nets and more compact decision trees. The “black-box” nature of neural networks was explained by extracting rules with both the domain expert and autoclass for both the continuous and the discrete valued inputs with rule sets performing better on the ‘low’ and ‘high’ levels. It follows from these analyses that performance at the two extremes was more important than average performance. It implied that the end user was particularly concerned with identifying mating with good potential and avoid mating with poor potential animals. The decision tree showed that when the herd performance was low then the foremost limiting factor was the dam performance whereas for medium and high herd performance sire level performance was the limiting factor. Through sensitivity analysis the most important and sensible factors with respect to productivity were sire breeding value and herd performance. It was, therefore, concluded that neural network rule extraction and decision tables were powerful management tools that allow the building of advanced and user-friendly decision-support systems for mating strategy designs and their evaluation.

An essential aspect of understanding any natural system is the ability to acquire knowledge through experience and to adapt to new situations. This study, investigates the use of back-propagation Artificial Neural Networks (ANN) (an artificial intelligence technique) approach to model and predict the performance of daughter first lactation milk yield in recorded dairy cattle herds in Kenya. Such prediction is a prerequisite to selection which ultimately leads to optimal breeding strategies and increased annual genetic progress. Data consisting of 6095 lactation records made by the Kenyan Holstein-Friesian cows from 76 officially milk recorded herds of 445 sires, 1956 dams and 2267 daughters and collected over the period 1988 -2005 were used to predict the first lactation performance of the female offspring based on recorded genetic traits of their parents using computer Neural Networks (NN). Weka and MATLAB softwares were used for NN analyses while SAS (2003) and Derivative free Restricted maximum likelihood (DFREML) (Meyer 1989) computer packages were used for statistical analyses. Different ANN were modeled and the best performing number of hidden layers and neurons and training algorithms retained. The performance of the ANN model in simulating daughter performance was compared with the industry default technique linear regression (LR) model. The best NN model had one hidden layer with 8 hidden nodes and tangent sigmoid transfer function for hidden layer. The correlation coefficients between the observed and the estimated daughter milk yield for the two estimation methods was generally high (>0.80). Including sire information resulted to more accurate predictions by a neural network as shown by reduced root mean square error. Generally more accurate prediction was obtained by a neural network approach than by linear regression. This suggests a non-linear relationship exists among the feature variables in the data and that these are learned by the hidden layer of the NN. Thus, prediction tests show that the ANN models used in this study

The effects of supplementing different diets with varying levels of energy and protein either in the morning or afternoon were assessed. 384 scavenging indigenous chickens aged 14 weeks were allocated to four treatment diets: 1: Positive control (commercial diet); 2: High energy (3062 kcal/kg DM), high protein (224 g/kg DM) (HEHP); 3: Low energy (2378 kcal/kg DM), high protein (218 g/kgDM) (LEHP); 4: Low energy, Low protein (LELP). The supplements were offered in the morning and afternoon. Results showed that both pre-lay scavenging pullets and adult hens that received supplemental diets with high protein levels (HEHP and LEHP) both in the morning and afternoon ate more supplemental feed during the afternoon hours. Supplemental energy intake by scavenging pullets was higher in the morning than in afternoon while the reverse was true for adult scavenging hens. The mean supplemental crude protein intake was higher for both pullets and hens in the afternoon. Scavenging indigenous pullets consumed 11.7% more supplemental crude protein in the afternoon while hens consumed 12.4% more supplemental crude protein in the afternoon. Supplemental lysine, tryptophan and methionine + cystine intakes were higher for scavenging pullets in the afternoon and the same case for adult scavenging hens in the morning and afternoon. Egg production and weights were higher for hens supplemented with HEHP and LEHP compared to those receiving commercial diet (CD), and those supplemented with LELP diets both in the morning and afternoon. Feed cost was higher in pre-lay pullets supplemented with commercial diets (same levels of energy and protein in morning and afternoon) than with HEHP and LEHP diets mainly due to the higher price of commercial feed compared to that of ingredients found locally. The study suggests that offering a supplement of commercial diet (same level of energy and protein) for growing scavenging pullets followed by a high energy and high protein supplement during the laying period may increase feed intake, nutrient intake, egg production and egg weights