The subject of clustering can be defined as the partitioning of a set of points in a multidimensional space into groups called clusters such that the points in each group are in some sense similar to one another . Finding these clusters is important because their points correspond to observations of different A kind of latent information that may be of interest are correlations in a data set . A correlation is a linear dependency between two or more features (attributes) of the data set . Amongst the dataset , there are points that have a kind of correlation; such as the points of a line or a plan . In this thesis we have tried to cluster a group of data based on a special kind . In this new approach , the points on a linear manifold situate into a cluster . Knowing about the existence of a relationship between features may enable us to learn hidden causalities . In this thesis we have tried to cluster a group of data based on a special kind . In this new approach , the points on a linear manifold situate into a cluster . Consider K as the maximum dimension for clusters to include . For k=1,..,K , we select a set of k+1 points from the whole dataset in a way that construct a set of k linearly independent vectors . Then , using the Gram–Schmidt process , we deduce an orthonormal basis for a k-dimensional manifold . Then we calculate the distances of the rest of points in the dataset from the center of this manifold , and make the distance histogram for the dataset . In the next step , we employ Kittler-Illingworth's method to find a minimum error threshold in the histogram . The points of the dataset are separated according to this threshold and the points with distances less than it are put into the k-dimension cluster . If this separation is good enough , the above procedure is repeated for this cluster with k=k+1 to find higher dimensional clusters , if any , until we cannot find any more clusters . Therefore in addition to construct new clusters , we can find clusters that exist in previous clusters . The process is restarted for the rest of the dataset with k=1 . Finally outliers are clustered in a single cluster , which clearly does form a manifold .