A Data-Centric Approach to Improve Machine Learning Model’s Performance in Production

Machine learning teaches computers to think in a similar way to how humans do. An ML models work by exploring data and identifying patterns with minimal human intervention. A supervised ML model learns by mapping an input to an output based on labeled examples of input-output (X, y) pairs. Moreover, an unsupervised ML model works by discovering patterns and information that was previously undetected from unlabelled data. As an ML project is an extensively iterative process, there is always a need to change the ML code/model and datasets. However, when an ML model achieves 70-75% of accuracy, then the code or algorithm most probably works fine. Nevertheless, in many cases, e.g., medical or spam detection models, 75% accuracy is too low to deploy in production. A medical model used in susceptible tasks such as detecting certain diseases must have an accuracy label of 98-99%. Furthermore, that is a big challenge to achieve. In that scenario, we may have a good working model, so a model-centric approach may not help much achieve the desired accuracy threshold. However, improving the dataset will improve the overall performance of the model. Improving the dataset does not always require bringing more and more data into the dataset. Improving the quality of the data by establishing a reasonable baseline level of performance, labeler consistency, error analysis, and performance auditing will thoroughly improve the model's accuracy. This review paper focuses on the data-centric approach to improve the performance of a production machine learning model.

PEMILIHAN METODE SEGMENTASI PADA CITRA ULTRASONOGRAFI OVARIUM

Penelitian ini membandingkan metode segmentasi untuk mengenali folikel pada citra ultrasonografi ovarium, metode segmentasi yang paling baik akan digunakan untuk proses perhitungan jumlah folikel. Penilaian kinerja metode segmentasi active contour dan active contour without edge dievaluasi menggunakan Probabilistic Rand Index (PRI) dan Global Consistency Error (GCE). Hasil penelitian ini menunjukkan metode segmentasi yang terbaikadalah active contour without edge karena memiliki nilai PRI lebih tinggi dan pada nilai GCE lebih rendah dari pada hasil metode segmentasi active contour.

Design and Implementation of a Real-Time Multi-Beam Sonar System Based on FPGA and DSP

Aiming at addressing the contradiction between the high-speed real-time positioning and multi-channel signal processing in multi-beam sonar systems, in this work we present a real-time multi-beam sonar system based on a Field Programmable Gate Array (FPGA) and Digital Signal Processing (DSP) from two perspectives, i.e., hardware implementation and software optimization. In terms of hardware, an efficient high-voltage pulse transmitting module and a multi-channel data acquisition module with time versus gain (TVG) compensation with characteristics such as low noise and high phase amplitude consistency, are proposed. In terms of algorithms, we study three beamforming methods, namely delay-and-sum (D&S), direct-method (DM) and Chirp Zeta Transform (CZT). We compare the computational efficiency of DM and CZT in the digital domain. In terms of software, according to the transmission bandwidth of the Gigabit Ethernet and a serial rapid IO (SRIO) interface, the data transmission paths of the acquired data and the beam pattern between the FPGA, the DSP, and a personal computer (PC) are planned. A master-slave multi-core pipelined signal processing architecture is designed based on DSP, which enhances the data throughput of the signal processor by seven times as compared with that of the single-core operation. The experimental results reveal that the sound source level of the transmitting module is around 190.25 dB, the transmitting beam width is 64° × 64°, the background noise of the acquisition module is less than 4 μVrms, the amplitude consistency error of each channel is less than −6.55 dB, and the phase consistency error is less than 0.2°. It is noteworthy that the beam number of the sonar system is 90 × 90, the scanning angle interval is 0.33°, the working distance ranges from 5 m to 40 m, and the maximum distance resolution is 0.384 m. In the positioning experiment performed in this work; the 3-D real-time position of the baffle placed in the detection sector is realized. Please note that the maximum deviation of azimuth is 2°, the maximum deviation of elevation is 2.3°, and the maximum distance deviation is 0.379 m.

EVALUASI TAMPILAN ANTARMUKA WEB SMK KOMPUTAMA MAJENANG MENGGUNAKAN METODE HEURIS

Web is a digital page that functions to process and display information. A website has become an important element for an educational institution to disseminate information more widely and be known to more people. A web with a bad interface can cause user to have trouble finding the information they need and does not want to browse or visit again. Therefore, the appearance of a good and friendly interface with user is very important to be implemented in order to make user feel at home and comfortable in accessing related information. Evaluation of the appearance of the SMK Komputama web interface is carried out to produce recommendations for improving the appearance of a better interface based on ease of use (usability) so that operational activities on the web can run well. Evaluation was carried out using the heuristic method by 3 respondents using a questionnaire based on aspects of the heuristic method. There are 10 aspects of the heuristic method used namely feedback, matching, freedom, consistency, error, efficiency, recognition, design, diagnosis, and documentation. The results showed that the appearance of the SMK Komputama web interface had been quite good in terms of ease of use (usability) but there were still some interface points that needed to be improved.

Tinjauan Desain Interface Website E-Commerce Wisata Mototravel.id Menggunakan Evaluasi Heuristik

This article is an overview of the website design interface of Lombok travel agencymorotravel.id. This review carried out using a heuristic evaluation introduced by Molich andNielsen that evaluates an interface in an information system. Heuristic evaluation is a guidethat can guide the design or used it as a tool for criticizing a decision taken. The aim is toimprove the model effectively. The evaluation includes visibility of system status, compatibilitybetween the system and the real world, control of user freedom, standards, and consistency,error prevention, user recognition of the system, flexibility and efficiency, aesthetic values andminimum values, system help, help features, and documentation. The purpose of thisevaluation is to measure morotravel.id's interface standard as one of the travel agent websitesin Lombok so that it can become a reference for information systems in the world of tourism.Based on the evaluation, it has stated that overall the morotravel.id site interface is feasible touse because it fulfills the completeness of the principles in the heuristic assessment. Still, ithas some things that need to be improved, especially in the help system and documentation.Heuristic evaluation is pragmatic and easy to do so that it gets quick results. Although thismethod does not produce high certainty, it is a relatively easy method to start an analysis ofinterface design. The hope for the next is the use of more than one way to improve the qualityof the analysis results because it meets the principle of complementarity in each evaluationmethod

Thresholding-based White Blood Cells Segmentation from Microscopic Blood Images

Digital image processing has a significant role in different research areas, including medical image processing, object detection, biometrics, information hiding, and image compression. Image segmentation, which is one of the most important steps in processing medical image, makes the objects inside images more meaningful. For example, from microscopic images, blood cancer can be identified which is known as leukemia; for this purpose at first, the white blood cells (WBCs) need to be segmented. This paper focuses on developing a segmentation technique for segmenting WBCs from microscopic blood images based on thresholding segmentation technique and it compares with the most commonly used segmentation technique which is known as color-k-means clustering. The comparison is done based on three well-known measurements, used for evaluating segmentation techniques which are probability random index, variance of information, and global consistency error. Experimental results demonstrate that the proposed thresholding-based segmentation technique provides better results compared to color-k-means clustering technique for segmenting WBCs as well as the time consumption of the proposed technique is less than the color-k-means which are 70.8144 ms and 204.7188 ms, respectively.

Optimal Cellular Automata Technique for Image Segmentation

Leukemia death secured 10 thplace among the most dangerous death in the world. The main reason is due to the delay in diagnosis which in turn delayed the treatment process. Hence it becomes an exigent requirement to diagnose leukemia in its early stage. Segmentation of WBC is the initial phase of leukemia detection using image processing.This paper aims to extract WBC from the image background. There exists various techniques for WBC segmentation in the literature. Yet, they provides inaccurate results.Cellular Automata can be effectively implemented in image processing. In this paper, we have proposed an Optimal Cellular Automata approach for image segmentation.In this approach, the optimal value for alive cells is obtained through particle swarm Optimization with Gravitational Search Algorithm (PSOGSA). The optimal value have fed in to the cellular automata model and get the segmented image. The results are validated based on the parameters likeRand Index (RI), Global Consistency Error (GCE), and Variation of Information (VOI). The Experimental results of proposed technique shows better results when compared to the previously proposed techniques namely, Hybrid K-Means with Cluster Center Estimation, Region Splitting and Clustering Technique and Cellular Automata. The proposed technique outperformed all other techniques.

Analysis of an observer strategy for initial state reconstruction of wave-like systems in unbounded domains

We are interested in reconstructing the initial condition of a wave equation in an unbounded domain configuration from measurements available in time on a subdomain. To solve this problem, we adopt an iterative strategy of reconstruction based on observers and time reversal adjoint formulations. We prove the convergence of our reconstruction algorithm with perfect measurements and its robustness to noise. Moreover, we develop a complete strategy to practically solve this problem on a bounded domain using artificial transparent boundary conditions to account for the exterior domain. Our work then demonstrates that the consistency error introduced by the use of approximate transparent boundary conditions is compensated by the stabilization properties obtained from the use of the available measurements, hence allowing to still be able to reconstruct the unknown initial condition.

Multilevel approximation of Gaussian random fields: Fast simulation

We propose and analyze several multilevel algorithms for the fast simulation of possibly nonstationary Gaussian random fields (GRFs) indexed, for example, by the closure of a bounded domain [Formula: see text] or, more generally, by a compact metric space [Formula: see text] such as a compact [Formula: see text]-manifold [Formula: see text]. A colored GRF [Formula: see text], admissible for our algorithms, solves the stochastic fractional-order equation [Formula: see text] for some [Formula: see text], where [Formula: see text] is a linear, local, second-order elliptic self-adjoint differential operator in divergence form and [Formula: see text] is white noise on [Formula: see text]. We thus consider GRFs on [Formula: see text] with covariance operators of the form [Formula: see text]. The proposed algorithms numerically approximate samples of [Formula: see text] on nested sequences [Formula: see text] of regular, simplicial partitions [Formula: see text] of [Formula: see text] and [Formula: see text], respectively. Work and memory to compute one approximate realization of the GRF [Formula: see text] on the triangulation [Formula: see text] of [Formula: see text] with consistency [Formula: see text], for some consistency order [Formula: see text], scale essentially linearly in [Formula: see text], independent of the possibly low regularity of the GRF. The algorithms are based on a sinc quadrature for an integral representation of (the application of) the negative fractional-order elliptic “coloring” operator [Formula: see text] to white noise [Formula: see text]. For the proposed numerical approximation, we prove bounds of the computational cost and the consistency error in various norms.