The current Ethiopian market is conducted in a traditional manner and market drivers are stillrnnot used for prediction of future market price. Although, large amount of market data have beenrngathered throughout years by both governmental and non-governmental organizations, yet littlernhave been done to analyze the data for future market price prediction. Moreover, the analysisrnmethods were often manual creating inefficiency in time and quality of market prediction.rnAnalyzing valuable data will show us what the future holds and accelerate the development goalsrnof the country in the sector. The study examines features of current Ethiopian market attributesrnto find out most valuable features for predicting market price. Eighteen technical indicators arerntaken and tested for their individual ability of prediction and redundancy. From the featurernselection of commodity marke, we have found that features like Stochastic %K, Stochastic %D,rnClose gain/loss, High, close price, Opening Price, Low, RSI, Ton and Moving AveragernConvergence/ divergence (MACD) founded to be in the top ten of individual performancernevaluation. Moreover features namely Stochastic %K, Relative Strength Index (RSI), BollingerrnBands-Upper, Highest-High, close gain/loss, Simple Moving Average (SMA), Closing price,rnMACD-Fast, Exponential Moving Average (EMA), MACD-Slow and Low founded to be lessrnredundant. The study also compares four machine learning models for their prediction ability ofrnEthiopian commodity market price. The outcomes of feature selection were used to compare thernmodels. Two experiments were conducted; the first was comparison of the models with 10 foldrncross validation using feature of high individual predictive ability and less redundancy. Thernsecond one was a comparison of models with separate train and test data using features of highrnindividual predictive ability and less redundancy. From the models (Support Vector Machinern(SVM), Artificial Neural Network (ANN), K-Nearest Neighbor (K-NN) and Ensemble Learning)rnthe performance of ANN and Ensemble Learning algorithms are shown to be accurate than SVMrnand K-NN. The average MAE rate of the ANN model was 2.8084. Ensemble Learning and SVMrnfollow with average MAE rate of 4.9362 and 8.1178 respectively. The other model was leastrnperformer with the MAE rate above 45.3381.