The minimum spanning tree clustering algorithm is known to be capable of detecting clusters with irregular boundaries. In this paper, we propose two based algorithms. first produces a k-partition set points for any given k. constructs the point and removes edges that satisfy predefined criterion. process repeated until k are produced. second partitions into group by maximizing overall standard deviation reduction, without value. We present our experimental results comparing proposed...

Explainable AI provides insights to users into the why for model predictions, offering potential better understand and trust a model, recognize correct predictions that are incorrect. Prior research on human explainable interactions has focused measures such as interpretability, trust, usability of explanation. There mixed findings whether can improve actual decision-making ability identify problems with underlying model. Using real datasets, we compare objective decision accuracy without...

A popular approach in current commercial anti-malware software detects malicious programs by searching the code of for scan strings that are byte sequences indicative code. The strings, also known as signatures existing malware, extracted malware analysts from samples, and stored a database often referred to virus dictionary. This process involves significant amount human efforts. In addition, there two major limitations this technique. First, not all have bit patterns evidence their nature....

Explainable AI provides insights to users into the why formodel predictions, offering potential for better un-derstand and trust a model, recognize correct AIpredictions that are incorrect. Prior research on human andexplainable interactions has typically focused measuressuch as interpretability, trust, usability of explanation.There mixed findings whether explainable can improveactual decision-making ability identify theproblems with underlying model. Using real datasets, wecompare objective...

Background Clear cell renal carcinoma (ccRCC) is the most common subtype of RCC, and accurate grading crucial for prognosis treatment selection. Biopsy reference standard grading, but MRI methods can improve complement procedure. Purpose Assess performance diffusion relaxation correlation spectroscopic imaging (DR‐CSI) in ccRCC. Study Type Prospective. Subjects 79 patients (age: 58.1 +/− 11.5 years; 55 male) with ccRCC confirmed by histopathology (grade 1, 7; grade 2, 45; 3, 18; 4, 9)...

Transparency has become a critical need in machine learning (ML) applications. Designing transparent ML models helps increase trust, ensure accountability, and scrutinize fairness. Some organizations may opt-out of transparency to protect individuals' privacy. Therefore, there is great demand for that consider both privacy security risks. Such can motivate improve their credibility by making the ML-based decision-making process comprehensible end-users. Differential (DP) provides an...

We propose two Euclidean minimum spanning tree based clustering algorithms — one a k-constrained, and the other an unconstrained algorithm. Our k-constrained algorithm produces k-partition of set points for any given k. The constructs representative removes edges that satisfy predefined criterion. process is repeated until k clusters are produced. partitions point into group by maximally reducing overall standard deviation in constructed from set, without prescribing number clusters. present...

Data mining techniques are playing an increasing role in making crucial decisions our daily lives ranging from credit card approvals to employment decisions. Typically algorithms used build decision models remain as a black-box the end user. Therefore process of appears be opaque. At same time, transparency model allows us discover hidden discrimination, and hold entities accountable. Although algorithmic with respect black box classifiers requires addressing many challenges, main objective...

Echo broadening effect (EBE) is inherent in three-dimensional range-gated imaging (3DRGI). The impacts the range-intensity profile of gate images which crucial three approaches 3DRGI based on depth scanning, supperresolution mapping and gain modulation. In this paper, we give space-time model EBE illustrates typical profiles under different temporal parameters laser pulse pulse. A head zone a tail exist both sides profiles. Our research demonstrates that should be suppressed scanning...

In today's society, AI systems are increasingly used to make critical decisions such as credit scoring and patient triage. However, great convenience brought by comes with troubling prevalence of bias against underrepresented groups. Mitigating in increase overall fairness has emerged an important challenge. Existing studies on mitigating focus eliminating sensitive demographic information embedded data. Given the temporal contextual complexity conceptualizing fairness, lossy treatment may...

AI Uncertainty Quantification (UQ) has the potential to improve human decision-making beyond predictions alone by providing additional probabilistic information users. The majority of past research on and concentrated model explainability interpretability, with little focus understanding impact UQ decision-making. We evaluated for instance-level UQ, calibrated using a strict scoring rule, in two online behavioral experiments. In first experiment, our results showed that was beneficial...

The desire of moving from data to intelligence has become a trend that pushes the world we live in today fast forward. Machine learning and mining techniques are being used as important tools unlock wealth voluminous amounts owned by organizations. Despite existing effort explaining their underlying machinery layman's terms, models output remain esoteric, discipline-based black boxes-viable only experts with years training development experiences. As gain growing popularity real world,...

Over the years social network data has been mined to predict individuals' traits such as intelligence and sexual orientation. While mining can provide many beneficial services user personalized experiences, it also harm when used in making critical decisions employment. In this work, we investigate reliability of applying techniques on various individual traits. spite preliminary success applications, paper, demonstrate vulnerabilities existing state art they are facing malicious attacks....

Aberrations in minimalist optical imaging systems pose significant challenges to achieving high-quality imaging. Traditional Wiener filtering methods, though effective, are constrained by their dependency on precise blur kernels and noise models, performance degrades with spatial variations these parameters. On the other hand, deep learning techniques often fail fully utilize prior information about aberrations suffer from limited interpretability. To address limitations, we propose a novel...

AI Uncertainty Quantification (UQ) has the potential to improve human decision-making beyond predictions alone by providing additional useful probabilistic information users. The majority of past research on and concentrated model explainability interpretability, with very little focus understanding impact UQ human-decision making performance. Moreover, is often equated model’s predicted probability, hence, not carefully calibrated. In this work, we implemented instance-based for three real...

Conventional wisdom holds that discrimination in machine learning is a result of historical discrimination: biased training data leads to models. We show the reality more nuanced; can be expected induce types bias not found data. In particular, if different groups have optimal models, and model for one group has higher accuracy, accuracy joint will disparate impact even when does display impact. argue due systemic bias, this likely situation, simply ensuring appears unbiased insufficient...

Aimed at improving the detection ability of night vision system, a new method to design optimum spectral band is presented. The uses MRC (Minimum Resolvable Contrast) as criteria system performance and brings spectrum characteristic scene into evaluation. According variety under different working bands, one which can maximize scent contrast doesn't obviously decrease cut-off spatial frequency curve. Finally by model, example special given.