Information Filtering via Improved Similarity Definition

Based on a new definition of user similarity, we introduce an improved collaborative filtering （ICF） algorithm, which couM improve the algorithmic accuracy and diversity simultaneously. In the ICF, instead of the standard Pearson coefficient, the user-user similarities are obtained by integrating the heat conduction and mass diffusion processes. The simulation results on a benchmark data set indicate that the corresponding algorithmic accuracy, measured by the ranking score, is improved by 6. 7% in the optimal case compared to the standard collaborative filtering （CF） algorithm. More importantly, the diversity of the recommendation lists is also improved by 63.6%. Since the user similarity is crucial for the CF algorithm, this work may shed some light on how to improve the algorithmic performance by giving accurate similarity measurement.     

Heterogeneity in initial resource configurations improves a network-based hybrid recommendation algorithm

Network-based recommendation algorithms for user–object link predictions have achieved significant developments in recent years. For bipartite graphs, the resource reallocation in such algorithms is analogous to heat spreading (HeatS) or probability spreading (ProbS) processes. The best algorithm to date is a hybrid of the HeatS and ProbS techniques with homogeneous initial resource configurations, which fulfills simultaneously high accuracy and large diversity requirements. We investigate the effect of heterogeneity in initial configurations on the HeatS + ProbS hybrid algorithm and find that both recommendation accuracy and diversity can be further improved in this new setting. Numerical experiments show that the improvement is robust.     

Temporal information effect on personalized recommendation

As information explosively grows, it requires to find out the highly efficient data from the global database. In this paper, based on two real systems, i.e., the MovieLens and Netflix, we investigate the temporal information effect on personalized recommendation for three collaborative filtering algorithms, by considering both the volume and time of the training data. It is observed that, for an increasing training data of a certain time range, the recommendation accuracy using the time-dependent training data outperforms that using the random training data of the same size. Moreover, the increase of data volume would not always improve the recommendation accuracy for the time-dependent training data case, whereas the recent temporal training data are important to elevate the recommendation accuracy. Further study on a hybrid algorithm of heat conduction and mass diffusion shows that, using the recent temporal information as training data, the algorithm reaches the optimal recommendation accuracy, when it returns to the original mass diffusion algorithm, which suggests no necessity of introducing an extra parameter. And the simple mass diffusion algorithm is found to be more advantageous than the previous three algorithms.     

Preference of online users and personalized recommendations

In a recent work [T. Zhou, Z. Kuscsik, J.-G. Liu, M. Medo, J.R. Wakeling, Y.-C. Zhang, Proc. Natl. Acad. Sci. 107 (2010) 4511], a personalized recommendation algorithm with high performance in both accuracy and diversity is proposed. This method is based on the hybridization of two single algorithms called probability spreading and heat conduction, which respectively are inclined to recommend popular and unpopular products. With a tunable parameter, an optimal balance between these two algorithms in system level is obtained. In this paper, we apply this hybrid method in individual level, namely each user has his/her own personalized hybrid parameter to adjust. Interestingly, we find that users are quite different in personalized hybrid parameters and the recommendation performance can be significantly improved if each user is assigned with his/her optimal personalized hybrid parameter. Furthermore, we find that users’ personalized parameters are negatively correlated with users’ degree but positively correlated with the average degree of the items collected by each user. With these understandings, we propose a strategy to assign users with suitable personalized parameters, which leads to a further improvement of the original hybrid method. Finally, our work highlights the importance of considering the heterogeneity of users in recommendation.     

Recommendation in evolving online networks

Recommender system is an effective tool to find the most relevant information for onlineusers. By analyzing the historical selection records of users, recommender system predictsthe most likely future links in the user-item network and accordingly constructs apersonalized recommendation list for each user. So far, the recommendation process ismostly investigated in static user-item networks. In this paper, we propose a model whichallows us to examine the performance of the state-of-the-art recommendation algorithms inevolving networks. We find that the recommendation accuracy in general decreases with timeif the evolution of the online network fully depends on the recommendation. Interestingly,some randomness in users’ choice can significantly improve the long-term accuracy of therecommendation algorithm. When a hybrid recommendation algorithm is applied, we find thatthe optimal parameter gradually shifts towards the diversity-favoring recommendationalgorithm, indicating that recommendation diversity is essential to keep a high long-termrecommendation accuracy. Finally, we confirm our conclusions by studying therecommendation on networks with the real evolution data.     

Personalized recommendation via an improved NBI algorithm and user influence model in a Microblog network

Bipartite network based recommendations have attracted extensive attentions in recent years. Differing from traditional object-oriented recommendations, the recommendation in a Microblog network has two crucial differences. One is high authority users or one’s special friends usually play a very active role in tweet-oriented recommendation. The other is that the object in a Microblog network corresponds to a set of tweets on same topic instead of an actual and single entity, e.g. goods or movies in traditional networks. Thus repeat recommendations of the tweets in one’s collected topics are indispensable. Therefore, this paper improves network based inference (NBI) algorithm by original link matrix and link weight on resource allocation processes. This paper finally proposes the Microblog recommendation model based on the factors of improved network based inference and user influence model. Adjusting the weights of these two factors could generate the best recommendation results in algorithm accuracy and recommendation personalization.     

An improved recommendation algorithm via weakening indirect linkage effect

We propose an indirect-link-weakened mass diffusion method(IMD), by considering the indirect linkage and the source object heterogeneity effect in the mass diffusion(MD) recommendation method. Experimental results on the MovieLens, Netflix, and RYM datasets show that, the IMD method greatly improves both the recommendation accuracy and diversity, compared with a heterogeneity-weakened MD method(HMD), which only considers the source object heterogeneity. Moreover, the recommendation accuracy of the cold objects is also better elevated in the IMD than the HMD method. It suggests that eliminating the redundancy induced by the indirect linkages could have a prominent effect on the recommendation efficiency in the MD method.     

Effects of high-order correlations on personalized recommendations for bipartite networks

In this paper, we introduce a modified collaborative filtering (MCF) algorithm, which has remarkably higher accuracy than the standard collaborative filtering. In the MCF, instead of the cosine similarity index, the user-user correlations are obtained by a diffusion process. Furthermore, by considering the second-order correlations, we design an effective algorithm that depresses the influence of mainstream preferences. Simulation results show that the algorithmic accuracy, measured by the average ranking score, is further improved by 20.45% and 33.25% in the optimal cases of MovieLens and Netflix data. More importantly, the optimal value depends approximately monotonously on the sparsity of the training set. Given a real system, we could estimate the optimal parameter according to the data sparsity, which makes this algorithm easy to be applied. In addition, two significant criteria of algorithmic performance, diversity and popularity, are also taken into account. Numerical results show that as the sparsity increases, the algorithm considering the second-order correlation can outperform the MCF simultaneously in all three criteria.     

改进Canny算法的红外人脸图像边缘轮廓提取

红外人脸图像的边缘轮廓特征对于红外人脸检测、识别等相关应用具有重要价值。针对红外人脸图像边缘轮廓提取时存在伪边缘的问题,提出了一种改进Canny算法的红外人脸图像边缘轮廓提取方法。首先通过对引导滤波算法引入“动态阈值约束因子”替换原始算法中的高斯滤波,解决了原始算法滤波处理不均匀和造成红外人脸图像弱边缘特征丢失的弊端;接着对原始算法的非极大值抑制进行了改进,在原始计算梯度方向的基础上又增加了4个梯度方向,使得非极大值抑制的插值较原始算法更加精细;最后改进OTSU(大津)算法,构造灰度-梯度映射函数确定最佳阈值,解决了原始算法人为经验确定阈值的局限性。实验结果表明:提出的改进Canny算法的红外人脸轮廓提取方法滤波后的图像,相较于原始Canny算法滤波处理,信噪比性能提升了34.40%,结构相似度性能提升了21.66%;最终的红外人脸边缘轮廓提取实验的优质系数值高于对比实验的其他方法,证明改进后的算法对于红外人脸图像边缘轮廓提取具有优越性。     

光学图像中的抛物线目标特征检测方法研究

针对光学图像中的抛物线目标特征检测问题,提出一种基于最小二乘法的抛物线检测方法。采用Canny边缘检测算子进行边缘提取,对提取到的边缘点采用最小二乘法确定抛物线方程的系数,通过对二次曲线施加抛物线约束,使得算法不需要经过迭代就能得到最优解,并对抛物线方程进行坐标变换推导,计算出抛物线的参数。仿真数据实验和实际图像的实验表明:该算法能够检测到光学图像中的抛物线目标特征,各个参数的计算精度均在98%以上。     

改进VGG网络的多聚焦图像的融合方法

为更好地保留原图像信息,提高图像融合性能,提出一种改进VGG卷积神经网络与边缘像素统计特征相结合的融合算法。首先,该算法将完整图像拆分成图像块,以图像块的预处理来获取较高的图像分类,精度达到0.985以上,以改进的VGG卷积神经网络来加快模型收敛速度,当图像块输入到网络当中,可以初步得到二分类的权值矩阵。其次,在高频细节部分,对于左聚焦图像和右聚焦图像的清晰模糊模块分别进行模糊化处理,根据像素点之间的统计特征经阈值分割后得到有明显边界的权值矩阵。最后,结合两次分割的权值矩阵,通过加权求和的融合策略,得到处处清晰的聚焦图像。为说明算法有效性,在实验部分展示其融合主观视觉效果图与信息熵等客观评价,该算法对比其他算法表现突出,可较好地保留原图像的信息。     

一类求解箱式约束优化问题的自适应引力搜索算法

为了改进引力搜索算法求解箱式约束优化问题的性能,提出了一类自适应引力搜索算法,新算法定义了算法停滞系数,当算法陷入停滞时,可以自适应的修改引力参数,帮助算法跳出停滞状态;定义了个体相似系数,当种群陷入局部最优时,通过变异策略改善种群的多样性。数值试验结果表明,新算法有效的平衡了全局开发和局部搜索能力,具有更强的全局寻优能力,适于求解复杂优化问题。     

ISG-FHEV等效燃油消耗最小控制策略

针对室内复杂环境下火灾识别准确率会降低的问题,提出了一种改进的粒子群算法优化支持向量机参数进行火灾火焰识别的方法。首先在 颜色空间进行火焰图像分割,对获得的火焰图像进行预处理并提取相关特征量;其次采用PSO算法搜索SVM的最优核参数和惩罚因子,并在PSO算法中加入变异操作和非线性动态调整惯性权值的方法,加快了搜索SVM最优参数的精度和速度;然后将提取的火焰各个特征量作为训练样本输入SVM模型进行训练,并建立参数优化后的SVM分类器模型;最后将待测试样本输入SVM模型进行分类识别。算法的火灾识别准确率达到94.09%,分类效果明显优于其他分类算法。仿真结果表明,改进的PSO优化SVM算法提高了火焰识别的准确率和实时性,算法的自适应性更强,误判率更低。     

改进的粒子群优化目标跟踪方法

针对粒子群优化算法应用在目标跟踪时,其惯性权重调节机制的局限性,提出了改进的粒子群优化目标跟踪方法。首先,对目标及粒子群算法中相应参数进行初始化;接着,引入粒子进化率的概念,对惯性权重调节机制进行改进,根据每代每个粒子的不同状态及时调整惯性权重;然后,在更新粒子的速度和位置的同时,更新个体最优解和全局最优解,进行下一次迭代;最后,比较粒子的适应度,选择相似性函数值最大的区域为目标。实验结果表明,该方法与使用自适应惯性权重调节机制的粒子群优化目标跟踪方法相比,减少了获取相同适应度所需的迭代次数,运算效率提高了42.9%。实现了目标在相似性函数出现"多峰"情况下的准确定位,对目标出现部分遮挡的情况具有很好的适应性。     

基于OTSU分割和HOG特征的行人检测与跟踪方法

传统的HOG算法针对整幅图像进行行人特征提取,大量的非人窗口计算必然降低检测的准确率和效率。为此,提出一种基于OTSU分割和HOG特征的行人检测与跟踪方法。利用OTSU算法以最佳阈值分割图像,在分割区域的基础上进行Canny边缘检测,通过边缘的对称性计算确定行人候选区,继而采用经PCA方法降维后的HOG特征和隐马尔可夫模型对行人候选区进行检测验证。最后,以确定的行人区域为跟踪窗口,利用CamShift算法跟踪行人。多组实验结果证明,本文方法的行人检测效率和精度均有所提高,跟踪性能稳定、可靠。     

基于改进Sobel算子的动态自动调焦算法研究

为了提高自动调焦算法在动态环境下的性能, 对调焦评价函数和调焦搜索策略进行了研究。在分析人类视觉系统特性研究成果的基础上,提出一种基于8方向Sobel算子边缘加权的调焦评价函数。同时,为了克服传统爬山法搜索速度慢的缺点,采用自适应变步长极值搜索策略,通过仿真实验可知,提出的8方向Sobel边缘检测算子具备良好的边缘提取效果,同时基于此算子结合人类视觉机制特性给予各边缘不同权重系数计算的调焦评价函数, 比经典两方向Sobel算子调焦评价函数具备更好的抗干扰能力。最后搭建基于液体透镜的实验平台,在动态环境下验证改进算法的性能。实验结果表明,提出的自动调焦算法在动态环境下调焦准确率达到97.5%。     

一种新的基于弧段提取的椭圆检测方法

为了克服椭圆检测过程中对椭圆完整性和边缘梯度信息依赖性较强的缺点,实现椭圆目标的快速有效检测,提出了一种新的基于弧段提取的椭圆拟合方法。首先将梯度方向符号相同的相邻边缘点连接成弧段,然后根据弧段的凸性和象限分类组成包含三段弧的弧组,定义了新的弧选择策略,利用位置约束和弧对的椭圆中心估计提取候选椭圆,最后采用改进的拟合算法拟合椭圆。实验结果表明,基于弧段提取的椭圆拟合方法相对于LMEDS算法和RHT3具有更好的准确性、鲁棒性和稳定性,实时性也有一定的提高。     

面向对象分析的非结构化背景目标高光谱探测方法研究

针对非结构化背景探测器中背景协方差矩阵估计的局限性,提出了一种基于面向对象分析的高光谱小目标探测算法。首先对图像进行自适应迭代分割处理,将其划分为许多均质对象;然后进行正态最优分布选取,利用多元正态无偏检验选取最佳对象集;最后将此数据集合作为局部背景并结合GLR基准算法进行目标探测。该算法可以使局部背景最大化的服从正态分布,有效地将背景光谱信息和目标光谱信息分离开来,同时通过最优选取过程克服了目标信息“污染”问题。为了验证算法的有效性,利用真实的OMIS数据进行仿真实验,并与非结构化背景探测器GLR和基于K-Means聚类的改进GLR算法的检测结果比较,结果表明提出的算法具有良好的探测性能和较低的虚警概率。     

改进RCF算法的电缆绝缘层边缘检测

目前电缆绝缘层厚度检测算法主要采用图像处理技术提取出绝缘层的边缘轮廓,此类算法存在绝缘层边缘过宽和边缘不连续等问题,影响了后续的检测精度.为提高绝缘层测量精度,新算法基于RCF算法进行改进,在模型的4、5阶段采用空洞卷积,增大模型的感受野;并在侧输出网络加入尺度增强模块(SEM模块)和由浅到深的级联网络,增加侧输出图像...     

一种边缘和过渡区域相结合的红外目标提取方法

传统的基于过渡区域提取的目标分割算法存在噪声敏感问题,从而会影响到过渡区域提取的准确性。与可见光图像相比,红外图像特别是红外光谱图像,受到探测器无法消除的热噪声影响,传统的目标提取算法准确率普遍降低。此外,虽然通过边缘能够精确定位目标,但是无法获取目标完整边缘。而过渡区域的灰度分布特点是可以解决基于边缘的目标提取难题。因此为了提高目标提取的抗噪性和准确性,提出了一种将过渡区域提取与边缘检测结合的自适应红外目标提取方法。首先利用像元空间邻域信息构造密度,以此有效降低噪声影响和获取图像边缘信息。然后基于像元密度信息最大分离目标边缘与背景,得到有效边缘和过渡区域,进而以此生长出目标。将边缘与过渡区域结合,可以很好地抑制噪声,多幅复杂场景实验评估了该方法的抗噪性能,结果显示,提出的方法在噪声的干扰下能较好的提取目标。