Deep learning for dermatologists: Part I. Fundamental concepts

β3 integrin immunohistochemistry as a method to predict sentinel lymph node status in patients with primary cutaneous melanoma

“…Analysis of digitized β3 integrin IHC slides relied on the well‐established H score concept 12 and was automated by a commercially available, reliable, and widely used software package. More contemporary and advanced machine learning approaches might have generated better results 16‐18 …”

How good is artificial intelligence (AI) at solving hairy problems? A review of AI applications in hair restoration and hair disorders

“…In supervised learning, algorithms learn how to perform tasks from examples with known outcomes (training). Then, the algorithm's ability to predict the correct outcome from an unknown example is evaluated (testing), and internal parameters are adjusted until the output is acceptable 5 . Classification or regression machine learning algorithms are used in medicine to predict discrete (eg, benign vs malignant) or continuous (eg, life expectancy) outcomes, respectively 4 .…”

“…Then, the algorithm's ability to predict the correct outcome from an unknown example is evaluated (testing), and internal parameters are adjusted until the output is acceptable. 5 Classification or regression machine learning algorithms are used in medicine to predict discrete (eg, benign vs malignant) or continuous (eg, life expectancy) outcomes, respectively. 4 The Framingham cardiovascular risk score is an example of a machine learning algorithm that was developed with significant input from clinicians and statisticians.…”

Government Policies Modeling in Controlling Indonesia's Covid-19 Cases Using Data Mining

“…Data mining classification techniques have been widely used in various fields of science, among which are often used is C4.5 algorithm such as diabetes prediction (Noviandi, 2018), heart disease prediction (Rohman, Suhartono, & Supriyanto, 2017), predictions of potential blood donors (Wahono & Riana, 2020), where C4.5 is a development of id3 algorithm. C4.5 can handle missing data, handling continuous data, pruning, rules, and also use the gain ratio as a solving criterion (Santosa & Ardian, 2018) In addition to the C4.5 algorithm, there is also a Deep Learning(DL) algorithm that has been widely used in Artificial Intelligence (AI), such as dermatology applications, where DL is a subset of machine learning, which can complete tasks or can answer specific questions (Murphree et al, 2020) In some studies that have been conducted by some researchers, the C4.5 algorithm provides considerable accuracy including 86.59% with an AUC value of 0.982 (Rohman et al, 2017), accuracy value of 70.32% (Noviandi, 2018), accuracy value of 90% (Bahri, Marisa Midyanti, Hidayati, Sistem Komputer, & Mipa, 2018)and accuracy value of 91.76% (Hijrah, Mukhlizar, & Pandria, 2020). However, some studies stated the C4.5 algorithm performs less well compared to other algorithms, and of them is DL (Mutrofin, Machfud, Satyareni, Ginardi, & Fatichah, 2020).…”