Study of distance metrics on k - nearest neighbor algorithm for star categorization

Classification of stars is essential to investigate the characteristics and behavior of stars. Performing classifications manually is error-prone and time-consuming. Machine learning provides a computerized solution to handle huge volumes of data with minimal human input. k-Nearest Neighbor (kNN) is one of the simplest supervised learning approaches in machine learning. This paper aims at studying and analyzing the performance of the kNN algorithm on the star dataset. In this paper, we have analyzed the accuracy of the kNN algorithm by considering various distance metrics and the range of k values. Minkowski, Euclidean, Manhattan, Chebyshev, Cosine, Jaccard, and Hamming distance were applied on kNN classifiers for different k values. It is observed that Cosine distance works better than the other distance metrics on star categorization.

Clustering-Based Plane Segmentation Neural Network for Urban Scene Modeling

Urban scene modeling is a challenging but essential task for various applications, such as 3D map generation, city digitization, and AR/VR/metaverse applications. To model man-made structures, such as roads and buildings, which are the major components in general urban scenes, we present a clustering-based plane segmentation neural network using 3D point clouds, called hybrid K-means plane segmentation (HKPS). The proposed method segments unorganized 3D point clouds into planes by training the neural network to estimate the appropriate number of planes in the point cloud based on hybrid K-means clustering. We consider both the Euclidean distance and cosine distance to cluster nearby points in the same direction for better plane segmentation results. Our network does not require any labeled information for training. We evaluated the proposed method using the Virtual KITTI dataset and showed that our method outperforms conventional methods in plane segmentation. Our code is publicly available.

A Decision Support System for Face Sketch Synthesis Using Deep Learning and Artificial Intelligence

The recent development in the area of IoT technologies is likely to be implemented extensively in the next decade. There is a great increase in the crime rate, and the handling officers are responsible for dealing with a broad range of cyber and Internet issues during investigation. IoT technologies are helpful in the identification of suspects, and few technologies are available that use IoT and deep learning together for face sketch synthesis. Convolutional neural networks (CNNs) and other constructs of deep learning have become major tools in recent approaches. A new-found architecture of the neural network is anticipated in this work. It is called Spiral-Net, which is a modified version of U-Net fto perform face sketch synthesis (the phase is known as the compiler network C here). Spiral-Net performs in combination with a pre-trained Vgg-19 network called the feature extractor F. It first identifies the top n matches from viewed sketches to a given photo. F is again used to formulate a feature map based on the cosine distance of a candidate sketch formed by C from the top n matches. A customized CNN configuration (called the discriminator D) then computes loss functions based on differences between the candidate sketch and the feature. Values of these loss functions alternately update C and F. The ensemble of these nets is trained and tested on selected datasets, including CUFS, CUFSF, and a part of the IIT photo–sketch dataset. Results of this modified U-Net are acquired by the legacy NLDA (1998) scheme of face recognition and its newer version, OpenBR (2013), which demonstrate an improvement of 5% compared with the current state of the art in its relevant domain.

A Blind Watermarking Model of the 3D Object and the Polygonal Mesh Objects for Securing Copyright

In this paper, we propose a novel model for 3D object watermarking. The proposed method is based on the properties of the discrete cosine transform (DCT) of the 3D object vertices to embed a secret grayscale image three times. The watermarking process takes place by using the vertices coefficients and the encrypted image pixels. Moreover, the extraction process is totally blind based on the reverse steps of the embedding process to recover the secret grayscale image. Various performance aspects of the method are measured and compared between the original 3D object and the watermarked one using Euclidean distance, Manhattan distance, cosine distance, and correlation distance. The obtained results show that the proposed model provides better performances in terms of execution time and invisibility.

Overview of attacks against civilian infrastructure during the Syrian civil war, 2012–2018

BackgroundHundreds of thousands of people have been killed during the Syrian civil war and millions more displaced along with an unconscionable amount of destroyed civilian infrastructure.MethodsWe aggregate attack data from Airwars, Physicians for Human Rights and the Safeguarding Health in Conflict Coalition/Insecurity Insight to provide a summary of attacks against civilian infrastructure during the years 2012–2018. Specifically, we explore relationships between date of attack, governorate, perpetrator and weapon for 2689 attacks against five civilian infrastructure classes: healthcare, private, public, school and unknown. Multiple correspondence analysis (MCA) via squared cosine distance, k-means clustering of the MCA row coordinates, binomial lasso classification and Cramer’s V coefficients are used to produce and investigate these correlations.ResultsFrequencies and proportions of attacks against the civilian infrastructure classes by year, governorate, perpetrator and weapon are presented. MCA results identify variation along the first two dimensions for the variables year, governorate, perpetrator and healthcare infrastructure in four topics of interest: (1) Syrian government attacks against healthcare infrastructure, (2) US-led Coalition offensives in Raqqa in 2017, (3) Russian violence in Aleppo in 2016 and (4) airstrikes on non-healthcare infrastructure. These topics of interest are supported by results of the k-means clustering, binomial lasso classification and Cramer’s V coefficients.DiscussionFindings suggest that violence against healthcare infrastructure correlates strongly with specific perpetrators. We hope that the results of this study provide researchers with valuable data and insights that can be used in future analyses to better understand the Syrian conflict.

Discretization of Fractional Operators: Analysis by Means of Advanced Computational Techniques

This paper studies the discretization of fractional operators by means of advanced clustering methods. The Grünwald–Letnikov fractional operator is approximated by series generated by the Euler, Tustin and generalized mean. The series for different fractional orders form the objects to be assessed. For this purpose, the several distances associated with the hierarchical clustering and multidimensional scaling computational techniques are tested. The Arc-cosine distance and the 3-dim multidimensional scaling produce good results. The visualization of the graphical representations allows a better understanding of the properties embedded in each type of approximation of the fractional operators.

EMR: A new metric to assess the resilience of directional mmWave channels to blockages

Millimeter-wave (mmWave) communication systems require narrow beams to compensate for high path loss and to increase the communication range. If an obstacle blocks the dominant communication direction, alternative paths (directions) should be quickly identified to maintain reliable connectivity. In this paper, we introduce a new metric to quantify the Effective Multipath Richness (EMR) of a directional communication channel in the angular domain. In particular, the proposed metric takes into account the strength and spatial diversity of the resolved Multipath Components (MPCs), while also considering the beamwidth of the communication link and the blockage characteristics. The metric is defined as a weighted sum of the number of distinct MPC clusters in the angular domain, where the clustering of the MPCs is performed based on the cosine-distance between the dominant MPCs. For a given transmitter (TX) and receiver (RX) pair, the EMR is a single scalar value that characterizes the robustness of the communication link against blockages, as it captures the number of unique communication directions that can be utilized. It is also possible to characterize the blockage robustness for the whole environment by evaluating the spatial distribution of the EMR metric considering various different TX/RX locations. Using our proposed metric, one can assess the scattering richness of different environments to achieve a particular service quality. We evaluate the proposed metric using our 28 GHz channel measurements in a library environment for Line-of-Sight (LOS) and NLOS scenarios, and compare it with some other commonly used propagation metrics. We argue that EMR is especially informative at higher frequencies, e.g., mmWave and terahertz (THz), where the propagation attenuation is high, and directional Non-Light-of-Sight (NLOS) communication is critical for the success of the network.

Clustering by Similarity of Brazilian Legal Documents Using Natural Language Processing Approaches

The Brazilian legal system postulates the expeditious resolution of judicial proceedings. However, legal courts are working under budgetary constraints and with reduced staff. As a way to face these restrictions, artificial intelligence (AI) has been tackling many complex problems in natural language processing (NLP). This work aims to detect the degree of similarity between judicial documents that can be achieved in the inference group using unsupervised learning, by applying three NLP techniques, namely term frequency-inverse document frequency (TF-IDF), Word2Vec CBoW, and Word2Vec Skip-gram, the last two being specialized with a Brazilian language corpus. We developed a template for grouping lawsuits, which is calculated based on the cosine distance between the elements of the group to its centroid. The Ordinary Appeal was chosen as a reference file since it triggers legal proceedings to follow to the higher court and because of the existence of a relevant contingent of lawsuits awaiting judgment. After the data-processing steps, documents had their content transformed into a vector representation, using the three NLP techniques. We notice that specialized word-embedding models—like Word2Vec—present better performance, making it possible to advance in the current state of the art in the area of NLP applied to the legal sector.

Divide-and-Conquer Federated Learning Under Data Heterogeneity

Federated Learning allows training of data stored in distributed devices without the need for centralizing training-data, thereby maintaining data-privacy. Addressing the ability to handle data heterogeneity (non-identical and independent distribution or non-IID) is a key enabler for the wider deployment of Federated Learning. In this paper, we propose a novel Divide-andConquer training methodology that enables the use of the popular FedAvg aggregation algorithm by over-coming the acknowledged FedAvg limitations in non-IID environments. We propose a novel use of Cosine-distance based Weight Divergence metric to determine the exact point where a Deep Learning network can be divided into class-agnostic initial layers and class-specific deep layers for performing a Divide and Conquer training. We show that the methodology achieves trained-model accuracy at-par with (and in certain cases exceeding) the numbers achieved by state-of-the-art algorithms like FedProx, FedMA, etc. Also, we show that this methodology leads to compute and/or bandwidth optimizations under certain documented conditions.