SUMMARY
Aliran masuk ke Danau Towuti, Kabupaten Luwu, Sulawesi Tengah, dianalisa dengan artificial neural network. Network mempunyai susunan yang terdiri dari 6 node pada layer input, 8 node pada hidden layer, dan 1 node pada layer output. Weight dari network dihitung dengan back propagation of error dan fungsi Sigmoid dipakai sebagai fungsi aktifasi. Dari analisa ditemukan bahwa curah aliran masuk (inflow) sangat dipengaruhioleh curah hujan di daerah-daerah Wawondula, Dam site, Timampu, Palumba, Loehadan Bantilang. Dari ke 6 derah tersebut, perubahan curah hujan di Wawondula mempunyai pengaruh yang sangat kuat terhadap inflowArtificial neural network method was applied in analyses of inflow to Towuti Lake of LuwuRegency, Central Sulawesi. The network was constructed by 6, 8, 1 nodes in input, hidden and output layers. The weights were calculated using back propagation of error and Sigmoid function was used as activation function. It was found that precipitations at Wawondula, Dam site, Timampu, Palumba, Loeha and Bantilang have the strongest influence to the inflow. Of the 6 areas, the precipitation changes at Wawondula has thestrongest effect to the change of the inflow