In the context of air-ground robot collaboration, the apperance of the moving ground object will change greatly from the perspective of the drone and traditional object tracking algorithms can hardly accomplish target tracking in such scenarios. In order to solve this problem, based on the Parallel Tracking And Detection (PTAD) framework and deep learning, an object detecting and tracking algorithm was proposed. Firstly, the Single Shot MultiBox detector (SSD) object detection algorithm based on Convolutional Neural Network (CNN) was used as the detector in the PTAD framework to process the keyframe to obtain the object information and provide it to the tracker. Secondly, the detector and tracker processed image frames in parallel and calculated the overlap between the detection and tracking results and the confidence level of the tracking results. Finally, the proposed algorithm determined whether the tracker or detector need to be updated according to the tracking or detection status, and realized real-time tracking of the object in image frames. Based on the comparison with the original algorithm of the PTAD on video sequences captured from the perspective of the drone, the experimental results show that the performance of the proposed algorithm is better than that of the best algorithm with the PTAD framework, its real-time performance is improved by 13%, verifying the effectiveness of the proposed algorithm.

The diversity between base classifiers and the accuracy of single base classifiers itself are two important factors that affect the generalization performance of ensemble system. Aiming at the problem that the diversity and accuracy are difficult to balance, a selective ensemble algorithm for gene expression data based on Diversity and Accuracy of Weighted Harmonic Average (D-A-WHA) was proposed. The Kernel Extreme Learning Machine (KELM) was used as the base classifier, and the diversity and accuracy of base classifiers were adjusted by D-A-WHA measure. Finally, a set of classifiers with high accuracy and high diversity with other base classifiers were selected to ensemble. The experimental results on UCI gene dataset show that compared with traditional Bagging, Adaboost and other ensemble algorithms, the classification accuracy and stability of the selective ensemble algorithm based on D-A-WHA measure are improved significantly,and it can be applied to the classification of cancer gene expression data effectively.

Since it was not ideal for Extreme Learning Machine (ELM) to deal with non-linear data, and the parameter randomization of ELM was not conducive for generalizing the model, an improved version of ELM algorithm was proposed. The parameters of ELM were optimized by Adaptive Chaotic Particle Swarm Optimization (ACPSO) algorithm to increase the stability of the algorithm and improve the accuracy of ELM for gene expression data classification. The simulation experiments were carried out on the UCI gene data. The results show that Adaptive Chaotic Particle Swarm Optimization-Extreme Learning Machine (ACPSO-ELM) has good stability and reliability, and effectively improves the accuracy of gene classification over existing algorithms, such as Detecting Particle Swarm Optimization-Extreme Learning Machine (DPSO-ELM) and Particle Swarm Optimization-Extreme Learning Machine (PSO-ELM).

Focusing on the issue that feature relevancy and dimension disaster problem in high-dimensional representation of Magnetic Resonance Imaging (MRI) prostate tumor Region of Interesting (ROI), a prostate tumor CAD model was proposed based on Neural Network (NN) with Principal Component Analysis (PCA) feature-level fusion in MRI. Firstly, 102 dimension features were extracted form MRI prostate tumor ROI, including 6 dimension geometry features, 6 dimension statistical features, 7 dimension Hu invariant moment features, 56 dimension GLCM texture features, 3 dimension Tamura texture features and 24 dimension frequency features. Secondly, 8 dimension features with cumulative contribution rate of 89.62% were obtained by using PCA in feature-level fusion, reducing the dimension of the feature vectors. Thirdly, the classical NN, which used Broyden-Fletcher-Goldfarb-Shanno (BFGS), Back-Propagation (BP) and Gradient Descent (GD), Levenberg-Marquardt as the training algorithm, was regarded as classifier to classify the features. Finally, 180 MRI images of prostate patients were used as original data, and the prostate tumor CAD model based on NN with feature-level fusion was utilized to diagnose. The experimental results illustrate that the ability to identify benign and malignant prostate tumor of neural network with PCA feature-level fusion is improved at least 10%, and the feature-level fusion strategy is effective, which increases the feature irrelevancy to a certain extent.

Noisy data and disease misjudgment in Region of Interest (ROI) of medical image is a typical inconsistent decision question of Inconsistent Decision System (IDS), and it is becoming huge challenge in clinical diagnosis. Focusing on this problem, based on certainty degree, inclusion degree and cover degree, a decision algorithm named ItoC-CIC was proposed for ROI of prostate tumor Magnetic Resonance Imaging (MRI) combined with macro-and-micro characteristics and global-and-local characteristics. Firstly, high-dimensional features for ROI of prostate tumor MRI were extracted to construct complete inconsistent decision table. Secondly, the equivalent classes possessing inconsistent samples were found by calculating certainty degree. Thirdly, the Score value was obtained by calculating inclusion degree and cover degree of inconsistent equivalent classes respectively, which was used to filter inconsistent samples, making inconsistent decision convert to consistent decision. Finally, test experiments of inconsistent decision tables were conducted on typical examples, UCI data and 102 features of MRI prostate tumor ROI. The experimental results illustrate that this algorithm is effective and feasible, and the conversion rate can reach 100% from inconsistent decision to consistent decision.

This paper is concerned with dynamic pricing problems of seasonal goods based on multi-Agent. The Q-learning algorithm and the Wolf-PHC (Win or Learn Fast, Policy Hill-Climbing) algorithm were proposed to learn the dynamic pricing model of seasonal goods which the two providers did not exchange information with each other. Finally, the paper obtained the simulation results of DF (Derivative Following) method, the Q-learning pricing algorithm and the Wolf-PHC pricing algorithm, and the compared results show that the Wolf-PHC pricing algorithm has a more effective optimization.

ased on 3ds MAX platform, simulation of the moving process of multiple aircrafts from various perspectives are realized in combination with 3D GIS by reading moving trace data (for example, data from black boxes), then a 3D simulation system for analyzing flying safety and accidents has been developed. This system could expedite the analysis of accidents and improve the accuracy of accident analysis.

The paper briefly introduced the concept of privacy preserving data mining technology and studied the application of decision tree classifier in this particular field. A decision tree classifier was applied and a scalar product protocol was added, so that the need of privacy preserving is satisfied as well as the advantage of decision tree is retained.