Random forest is an ensemble classification algorithm. It performs well when most predictive variables are noisy and can be used when the number of variables is much larger than the number of observations. The use of bootstrap samples and restricted subsets of attributes makes it more powerful than simple ensembles of trees. The main advantage of a random forest classifier is its explanatory po...

the needs of Internet applications QoS guarantee increased the demand of internet traffic classification, especially for interactive real time applications. Therefore, several classification methods were developed. Machine Learning (ML) classification is one of the most modern techniques, which solves the problem of traditional port base method. This paper compared experimentally the accuracy o...

Well-designed visualizations have an important role to play to aid in the public’s understanding of algorithms. This work presents a set of design principles for using visualization to explain machine learning algorithms specifically, and demonstrates these principles applied to the operations of the random forest algorithm.

We report a scaling of the mean matter density with the width of the saturated Lyα absorptions. This property is established using the “pseudo-hydro” technique (Croft et al. 1998). It provides a constraint for the inversion of the Lyα forest, which encounters difficulty in the saturated region. With a Gaussian density profile and the scaling relation, a simple inversion of the simulated Lyα for...

Data mining is the process of extracting and analyzing the large datasets to find out various hidden relationship patterns and much other useful information. Random forest is an ensemble method which is widely used is application having large datasets because of its interesting features like handling imbalanced data, identifying variable importance and detecting error rate. For building random ...

Conventional n-best reranking techniques often suffer from the limited scope of the n-best list, which rules out many potentially good alternatives. We instead propose forest reranking, a method that reranks the packed forest of exponentially many parses. Although exact inference is intractable with non-local features, we present an approximation algorithm that makes discriminative training pra...

Abstract To study the structure of solutions for random k-SAT and random CSPs, this paper introduces the concept of average similarity degree to characterize how solutions are similar to each other. It is proved that under certain conditions, as r (i.e. the ratio of constraints to variables) increases, the limit of average similarity degree when the number of variables approaches infinity exhib...

We study the problem of uniformly partitioning the edge set of a tree with n edges into k connected components, where k ≤ n. The objective is to minimize the ratio of the maximum to the minimum number of edges of the subgraphs in the partition. We show that, for any tree and k ≤ 4, there exists a k-split with ratio at most two. (Proofs for k = 3 and k = 4 are omitted here.) For general k, we pr...

Player movements in team sports are often complex and highly correlated with both nearby and distant players. A single motion model would require many degrees of freedom to represent the full motion diversity of each player and could be difficult to use in practice. Instead, we introduce a set of Game Context Features extracted from noisy detection data to describe the current state of the matc...