A key challenge for maximizing the world’s food supply is crop yield prediction. In this study, three machine models are used to predict the fresh weight (yield) of lettuce plants that are grown inside indoor farms hydroponically using the vertical farming infrastructure, namely, support vector regressor (SVR), random forest regressor (RFR), and deep neural network (DNN).The climate data, nutrient data, and plant growth data are passed as input to train the models to understand the growth pattern based on the available features. The study of evaluation metrics majorly covers Mean Squared Error (MSE), Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), R-squared, and Adjusted R-squared values.The results of the project have shown that the Random Forest with all the features is the best model having the best results with the least cross-validated MAE score and good cross-validated Adjusted R-squared value considering that the error of the prediction is minimal. This is followed by the DNN model with minor differences in the resulting values. The Support Vector Regressor (SVR) model gave a very poor performance with a huge error value that cannot be afforded in this scenario. In this study, we have also compared various evaluating metrics mentioned above and considered the cross-validated MAE and cross-validated Adjusted R-squared metrics. According to our study, MAE had the lowest error value, which is less sensitive to the outliers and adjusted R-squared value helps to understand the variance of the target variable with the predictor variable and adjust the metric to prevent the issues of overfitting.

Made available in DSpace on 2015-03-26T13:32:05Z (GMT). No. of bitstreams: 1 texto completo.pdf: 688077 bytes, checksum: 369ec0145d58b4c3f2d93ab69403df95 (MD5) Previous issue date: 2009-01-19 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / Many times, the interpretation of the results in cluster analysis is done subjectively, that is, through inspection on dendograms, since there are no objective criteria to identify the formed clusters. In face of such a problem, the present study aimed to: (1) find out an objective way to achieve the cut-point (optimal number of clusters) in a dendogram in order to help on taking the right decision; (2) work out index concepts such as Root Mean Square Standard Deviation (RMSSTD) and R Squared (RS), explaining the contribution of each one of them in determining the optimal number of cluster; (3) method application, aiming to identify divergent accessions that will be used on improvement programs. An alternative solution for this problem is to use the RMSSTD and RS which are calculated according to the number of variables among and within the clusters formed, characterizing an objective way to determine the optimal number. Another solution is achieved by using the RS. Some morphological characteristics of the forty nine accessions of the species Capsicum chinense Jacq. from the Germplasm Bank of Vegetables of the Federal University of Viçosa (Banco de Germoplasma de Hortaliças da Universidade Federal de Viçosa, Minas Gerais Brazil) were analyzed by means of cluster analysis. The accessions were clustered based on the proposed techniques and an optimal number of clusters was achieved. The 49 accessions analyzed were classified into only seven clusters according to the graph of the RMSSTD versus the number of clusters and the graph of the RS versus the number of clusters. / Muitas vezes, a interpretação dos resultados em análise de agrupamentos é feita de forma subjetiva, isto é, através da inspeção de dendrogramas. Isto se deve ao fato de haver dificuldade em se encontrar na literatura um critério objetivo de fácil aplicação para identificar o número ideal de grupos formados. Diante deste problema, o presente trabalho teve por objetivos: 1) Avaliar a aplicabilidade de critério objetivo de se obter o ponto de corte (número ótimo de clusters) num dendrograma para a tomada de decisão; 2) trabalhar os conceitos de índices como RMSSTD (root mean square standard deviation) e RS (R-Squared), discutindo a contribuição de cada um destes na obtenção do número ótimo de clusters em acessos de Capsicum chinense; 3) aplicação do método, visando a identificar acessos divergentes de Capsicum chinense para serem utilizados em programas de melhoramento. Os índices RMSSTD e RS são calculados de acordo com as variáveis entre e dentro dos grupos formados, caracterizando uma forma objetiva para determinar o número ótimo. Para se obter o ponto de máxima curvatura da trajetória dos índices RMSSTD e RS em função do aumento do número de grupos (X), utilizou-se o Método da Máxima Curvatura Modificado. Foram analisadas, por meio da análise de agrupamentos, algumas características morfológicas de quarenta e nove acessos da espécie Capsicum chinense Jacq. do Banco de Germoplasma de Hortaliças da Universidade Federal de Viçosa. A partir das técnicas propostas agrupou-se os acessos, obtendo um número ótimo de grupos. Os resultados classificam os 49 acessos avaliados em apenas sete grupos de acordo com o gráfico do RMSSTD versus o número de grupos e o gráfico do RS versus o número de grupos.