This paper addresses the challenge of dimensionality reduction (DR) and clustering, which are often treated separately in traditional learning approaches. It proposes a joint approach that optimizes both tasks simultaneously, aiming to improve the performance of both. The authors assume that the transformation from the latent space to the data is nonlinear and unknown. To recover 'clustering-friendly' latent representations, they propose a deep neural network (DNN) for DR and a K-means clustering algorithm. The DNN is designed to approximate any nonlinear function, making the approach suitable for a broad class of generative models. The paper includes a detailed design of the DNN structure and the optimization criterion, along with an effective and scalable algorithm for solving the formulated optimization problem. Experiments on various real datasets demonstrate the effectiveness of the proposed method, showing significant improvements over state-of-the-art techniques. The code for the experiments is available online.This paper addresses the challenge of dimensionality reduction (DR) and clustering, which are often treated separately in traditional learning approaches. It proposes a joint approach that optimizes both tasks simultaneously, aiming to improve the performance of both. The authors assume that the transformation from the latent space to the data is nonlinear and unknown. To recover 'clustering-friendly' latent representations, they propose a deep neural network (DNN) for DR and a K-means clustering algorithm. The DNN is designed to approximate any nonlinear function, making the approach suitable for a broad class of generative models. The paper includes a detailed design of the DNN structure and the optimization criterion, along with an effective and scalable algorithm for solving the formulated optimization problem. Experiments on various real datasets demonstrate the effectiveness of the proposed method, showing significant improvements over state-of-the-art techniques. The code for the experiments is available online.