Nowadays, clustering the stock market is a problem because it is difficult to find the series with similarities in patterns, regularities, and outliers due to its non-linear and dynamic nature. Analytical methods for such situations include K-means utilizing Euclidean distance and dynamic time warping to manage linear and non-linear sequence alignment distances in time series data. Both methods were employed to cluster the time point and are still limited in clustering the series. We present a different approach using a model-based generalization of the Clusterwise Vector Autoregressive algorithm, using the integration of the neural network model and K-means clustering which we call Clusterwise Neural Network. For this analysis, stock price information from companies operating in the financial sector was utilized from July 1, 2023 until June 30, 2024. The stock price variables utilized were opening, closing, and high. The empirical results reveal that the Clusterwise Neural Network technique demonstrated superior performance in clustering stock prices within the financial sector, as evidenced by a higher silhouette score.