基于K-means-LSTM模型的证券股价预测

鉴于股票数据具有非平稳、非线性等特征,传统的统计模型无法精准预测股票价格的未来趋势。针对这个问题,构建一种混合深度学习方法来提高股票预测性能。首先,通过将距离算法修改为DTW(动态时间归整),令K-means聚类算法拓展为更适用于时间序列数据的K-means-DTW,聚类出价格趋势相似的证券;然后,通过聚类数据来训练LSTM(长短时记忆网络)模型,以实现对单支股票价格的预测。实验结果表明,混合模型K-means-LSTM表现出更好的预测性能,其预测精度和稳定性均优于单一LSTM模型。

K均值聚类在葡萄酒分级中的应用

葡萄酒分级是葡萄酒评价中的一个重要内容,结合数据降维技术,建立葡萄酒K均值分类模型。通过实例分析证实所得结果较好,该方法在葡萄酒的评价分级中具有很好的应用价值。

在线约会对象推荐模型

限制选择集的尺寸,往往会得到更好的结果。针对于网络约会而言,给予用户少量且多样的约会对象推荐,有利于网恋成功率的提高。首先根据采集的用户信息进行分类,包括表面信息:照片、姓名、身高、体重、年龄、月收入、住址、是否接受异地恋、ELO评分、择偶标准;深度信息:学历、性格;多样性信息:爱好;然后构建K均值聚类模型和协同过滤模型相结合的在线约会对象推荐模型,考虑到推荐对象多样性,所以先利用K均值聚类模型将所有用户分类并判断待匹配用户所属类别,然后,考虑推荐对象深度,所以运用协同过滤模型分别计算不同信息特征下,待匹配用户与其他用户之间的相似度,将三种特征相似度赋权重为0.2,、0.4、0.4,得到最后的总相似度。最后根据不同用户特征信息不同,得到不同的选择集尺寸,根据多次模拟得到临界相似度,具体原则为:小于临界相似度的排除,在待匹配用户所属类别中选择5个,其他类别中各选1个。通过此种方法便可得到最优的约会对象集。

基于Python的水下导航适配区分类预测研究

在探讨水下导航系统的区域适配性标定问题时,本研究首先对所提供的重力异常值数据集执行了插值算法,以增强基准图的分辨率。随后,采用Python编程语言实现的k-means聚类算法对数据进行空间分割,并对各个子区域进行精确标定。通过对标定结果进行编码,并选取与研究目标密切相关的13个关键指标,运用主成分分析(PCA)方法进行降维处理,以简化模型复杂度并提取最具代表性的特征。进一步构建了逻辑回归模型,通过两次迭代优化,提高分类准确性。通过将模型预测结果与实际值进行比较,构建了接收者操作特征(ROC)曲线,以评估模型的预测性能。通过与标准编码的比较,验证了模型在预测分类区域适配性方面的有效性。在模型迁移性预测方面,对新数据集执行了相同的预处理流程,并在此基础上对仿真参数进行了调整,具体包括上下5%和10%的变动。通过灵敏度分析,绘制了参数变化与模型准确率之间的关系图,从而深入探讨了模型参数对预测结果的影响,进一步验证了模型的鲁棒性和适用性。综合分析结果表明,在推动“海洋强省”建设的战略背景下,实现海洋经济发展规划的关键之一在于海洋高新技术领域的创新。其中,水下导航与定位技术的适配区分类预测技术是核心技术之一。水下航行器在执行任务时,需确保自主性、无源性、高隐蔽性、不受地域和时间限制以及高精度的导航与定位能力。重力辅助导航技术是实现上述要求的有效方法之一。本研究的成果为水下导航系统的适配性标定提供了科学的方法论和技术支持,对于提升水下航行器的导航与定位能力具有重要意义。When exploring the regional adaptability calibration issue of underwater navigation systems, this study first performed interpolation algorithms on the provided gravity anomaly value dataset to enhance the resolution of the reference map. Subsequently, the k-means clustering algorithm, implemented in the Python programming language, was used to spatially segment the data and precisely calibrate each sub-region. The calibration results were encoded, and 13 key indicators closely related to the research objectives were selected for dimensionality reduction using Principal Component Analysis (PCA) to simplify model complexity and extract the most representative features. A logistic regression model was further constructed, and its classification accuracy was improved through two iterations of optimization. By comparing the model’s predicted results with actual values, a Receiver Operating Characteristic (ROC) curve was constructed to assess the model’s predictive performance. The effectiveness of the model in predicting regional adaptability was verified by comparing it with standard encoding. In terms of model translatability prediction, the same preprocessing procedures were performed on a new dataset, and simulation parameters were adjusted accordingly, including variations of 5% and 10% up and down. Through sensitivity analysis, a relationship diagram between parameter changes and model accuracy was plotted, thereby deeply exploring the impact of model parameters on prediction results and further verifying the model’s robustness and applicability. The comprehensive analysis results indicate that one of the keys to promoting the construction of a “Marine Strong Province” under the strategic background of ocean economic development planning lies in innovation in the field of marine high-tech. Among them, the classification prediction technology of adaptive areas for underwater navigation and positioning technology is one of the core technologies. Underwater vehicles need to ensure autonomy, passivity, high concealment, unrestricted by geography and time, and high-precision navigation and positioning capabilities when performing tasks. Gravity-assisted navigation technology is one of the effective methods to achieve the above requirements. The results of this study provide a scientific methodology and technical support for the adaptability calibration of underwater navigation systems, which is of great significance for enhancing the navigation and positioning capabilities of underwater vehicles.

海洋数据下的密度自适应聚类算法

针对DBSCAN算法需要人工设定参数,且数在对不同疏密度的数据敏感度较低以及处理多维多密度的海洋数据时鲁棒性欠佳的问题,提出一种基于K-均值模型的多密度自适应聚类算法AM-DBSCAN(adaptive multi-density DBSCAN algorithm)。采用K-均值模型对数据进行初次聚类,分别以结果簇中距离最远两点的平均值及最小簇的样本数作为DBSCAN算法中的邻域半径(Eps)及邻域样本阈值(Minpts);以最短路径原则改进DBSCAN算法中Eps邻域判定方式,提高算法全局的可靠性及稳定性。实验结果表明,相对于DBSCAN聚类算法,AM-DBSCAN算法在处理密度不均的数据时在聚类准确度和聚类效率方面有所提升。

Bag-of-visual-words model for artificial pornographic images recognition

It is illegal to spread and transmit pornographic images over internet,either in real or in artificial format.The traditional methods are designed to identify real pornographic images and they are less efficient in dealing with artificial images.Therefore,criminals turn to release artificial pornographic images in some specific scenes,e.g.,in social networks.To efficiently identify artificial pornographic images,a novel bag-of-visual-words based approach is proposed in the work.In the bag-of-words(Bo W)framework,speeded-up robust feature(SURF)is adopted for feature extraction at first,then a visual vocabulary is constructed through K-means clustering and images are represented by an improved Bo W encoding method,and finally the visual words are fed into a learning machine for training and classification.Different from the traditional BoW method,the proposed method sets a weight on each visual word according to the number of features that each cluster contains.Moreover,a non-binary encoding method and cross-matching strategy are utilized to improve the discriminative power of the visual words.Experimental results indicate that the proposed method outperforms the traditional method.

THRFuzzy:Tangential holoentropy-enabled rough fuzzy classifier to classification of evolving data streams

The rapid developments in the fields of telecommunication, sensor data, financial applications, analyzing of data streams, and so on, increase the rate of data arrival, among which the data mining technique is considered a vital process. The data analysis process consists of different tasks, among which the data stream classification approaches face more challenges than the other commonly used techniques. Even though the classification is a continuous process, it requires a design that can adapt the classification model so as to adjust the concept change or the boundary change between the classes. Hence, we design a novel fuzzy classifier known as THRFuzzy to classify new incoming data streams. Rough set theory along with tangential holoentropy function helps in the designing the dynamic classification model. The classification approach uses kernel fuzzy c-means(FCM) clustering for the generation of the rules and tangential holoentropy function to update the membership function. The performance of the proposed THRFuzzy method is verified using three datasets, namely skin segmentation, localization, and breast cancer datasets, and the evaluated metrics, accuracy and time, comparing its performance with HRFuzzy and adaptive k-NN classifiers. The experimental results conclude that THRFuzzy classifier shows better classification results providing a maximum accuracy consuming a minimal time than the existing classifiers.