<![CDATA[Accurate crop yield prediction is essential to support agricultural planning, food supply stability, and decision-making in modern precision agriculture. Agricultural production is influenced by many complex and nonlinear factors such as rainfall, temperature, humidity, soil conditions, and fertilizer usage. Traditional statistical methods often face limitations in handling high-dimensional and nonlinear agricultural datasets. Therefore, this study proposes a crop yield prediction model using Artificial Neural Network (ANN) combined with Principal Component Analysis (PCA) for dimensionality reduction. PCA is applied in the preprocessing stage to reduce redundant and correlated input variables while preserving the most important data variance. The reduced dataset is then used to train the ANN model to predict crop yield values. The model is implemented using Python with libraries including NumPy, Pandas, Scikit-learn, and TensorFlow/Keras. The dataset used in this research consists of 1000 agricultural records covering three crop commodities, namely maize, barley, and rice. Model performance is evaluated using visualization techniques including histogram error, histogram predicted, PCA explained variance, predicted vs actual plot, residual plot, and training history graph. Experimental results show that the PCA-ANN model produces accurate and stable prediction results with low prediction error and strong agreement between predicted and actual values. The integration of PCA and ANN improves prediction performance, reduces computational complexity, and minimizes overfitting risk. This research demonstrates that the PCA-ANN approach is effective for crop yield prediction and can support data-driven agricultural decision-making.]]>
