scholarly journals Integrating Learning from Examples into the Search for Diagnostic Policies

2005 ◽  
Vol 24 ◽  
pp. 263-303 ◽  
Author(s):  
V. Bayer-Zubek ◽  
T. G. Dietterich
Keyword(s):  
Decision Making ◽  
Greedy Algorithms ◽  
Search Space ◽  
Search Algorithms ◽  
Systematic Search ◽  
Training Data ◽  
Test Results ◽  
New Family ◽  
Diagnostic Decision ◽  
Training Examples

This paper studies the problem of learning diagnostic policies from training examples. A diagnostic policy is a complete description of the decision-making actions of a diagnostician (i.e., tests followed by a diagnostic decision) for all possible combinations of test results. An optimal diagnostic policy is one that minimizes the expected total cost, which is the sum of measurement costs and misdiagnosis costs. In most diagnostic settings, there is a tradeoff between these two kinds of costs. This paper formalizes diagnostic decision making as a Markov Decision Process (MDP). The paper introduces a new family of systematic search algorithms based on the AO* algorithm to solve this MDP. To make AO* efficient, the paper describes an admissible heuristic that enables AO* to prune large parts of the search space. The paper also introduces several greedy algorithms including some improvements over previously-published methods. The paper then addresses the question of learning diagnostic policies from examples. When the probabilities of diseases and test results are computed from training data, there is a great danger of overfitting. To reduce overfitting, regularizers are integrated into the search algorithms. Finally, the paper compares the proposed methods on five benchmark diagnostic data sets. The studies show that in most cases the systematic search methods produce better diagnostic policies than the greedy methods. In addition, the studies show that for training sets of realistic size, the systematic search algorithms are practical on today's desktop computers.

Transformer + GAN based Traditional Chinese Medicine inpatient prescription recommendation (Preprint)

2021 ◽  
Author(s):  
Hong Zhang
Keyword(s):  
Decision Making ◽  
Chinese Medicine ◽  
Traditional Chinese Medicine ◽  
Electronic Medical Records ◽  
Clinical Diagnosis ◽  
Medical Records ◽  
Training Data ◽  
Diagnosis And Treatment ◽  
Test Results ◽  
Treatment Scheme

BACKGROUND Clinical diagnosis and treatment decision making support is at the core of medical artificial intelligent research, in which Traditional Chinese Medicine (TCM) decision making is an important part. Traditional Chinese Medicine is a traditional medical system originated from China, of which the main clinical model is to conduct individualized diagnosis and treatment by relying on the four-diagnosis information. One of the key tasks of the TCM artificial intelligence research is to develop techniques and methods of clinical prescription decision making which takes all the relevant information of a patient as input, and produces a diagnosis and treatment scheme as output. Given the complexity of TCM clinical diagnosis and treatment schemes, decision making support of clinical diagnosis and treatment schemes remains as a research challenge for lacking of an effective solution. Fortunately, as the volume of the massive clinical data in the form of electronic medical records increases rapidly, it becomes possible for the computer to produce personalized diagnosis and treatment scheme recommendation through machine learning on the basis of the clinical big data. OBJECTIVE The objective of this research is to develop a real-time diagnosis and treatment scheme recommendation model for TCM inpatients. This is accomplished by using historical clinical medical records as training data to train a Transformer network. Furthermore, to alleviate the issue of overfitting, a Generative Adversarial Network is used to generate noise-added samples from the original training data. These noise-added samples along with the original samples form the complete train data set. METHODS valid information, such as the patient’s current sickness situation, medicines taken, nursing care given, vital signs, examinations and test results, is extracted from the patient’s electronic medical records, then the obtained information is sorted chronically, to produce a sequence of data of each patient. These time-sequence data is then used as input to the Transformer network. The output of the network would be the prescription information a physician would give. Overfitting is a common problem in machine learning, and becomes especially server when the network is complex with insufficient training data. In this research, a Generative Adversarial Network, is used to double the number of training samples by producing noise-added samples from the original samples. This, to a great extent, lessens the overfitting problem. RESULTS A total of 21,295 copies of inpatient electronic medical records from Guang’anmen traditional Chinese medicine hospital was used in this research. These records were created between January 2017 and December 2018, covering a total of 6352 kinds of medicines. These medicines were sorted into 829 types of first category medicines based on the class relationships among medicines. As shown by the test results, the performance of a fully trained Transformer model can have an average precision rate of 80.58%，and an average recall rate of 68.49%. CONCLUSIONS As shown by the preliminary test results, the Transformer-based TCM prescription recommendation model outperforms the existing conventional methods. The extra training samples generated by the GAN network helps to overcome the overfitting issue, leading a further improved recall rate and precision rate.

Prediction of User Loyalty Using the Naive Bayes Method in the "Goprint" Online Printing Marketplace

2020 ◽  
Vol 8 (3) ◽  
pp. 227
Author(s):  
Gede Widiastawan ◽  
I Gusti Agung Gede Arya Kadyanan
Keyword(s):  
Decision Making ◽  
Naive Bayes ◽  
Naïve Bayes ◽  
Training Data ◽  
Test Results ◽  
Bayes Method ◽  
Testing Data ◽  
Bayes Algorithm ◽  
Naive Bayes Method

Goprint is an Online Printing Marketplace that connects printing services with users who want to print documents quickly without the need to queue. In the span of time from April 2019 to September 2019 it was found that the number of Goprint users reached 407 users, 24 partners, and 256 orders. From transactions that have been carried out by users, not a few orders are often canceled due to ineffective Goprint features or poor partner performance. This causes Goprint users to feel dissatisfied with the services provided by the Goprint application. The Naive Bayes algorithm is one of the algorithms used for classification or grouping of data, but can also be used for decision making. With this algorithm and the problems that occur, the authors make a system to predict the loyalty of Goprint users to anticipate users who stop leaving Goprint because they are not satisfied or loyal users. The data used as training data is 20 and testing data is 10. From the test results it is found that the value of precision is 80%, 100% recall, and 90% accuracy.

scholarly journals Deep Learning-Based Differentiation between Mucinous Cystic Neoplasm and Serous Cystic Neoplasm in the Pancreas Using Endoscopic Ultrasonography

Diagnostics ◽  
2021 ◽  
Vol 11 (6) ◽  
pp. 1052
Author(s):  
Leang Sim Nguon ◽  
Kangwon Seo ◽  
Jung-Hyun Lim ◽  
Tae-Jun Song ◽  
Sung-Hyun Cho ◽  
...  
Keyword(s):  
Decision Making ◽  
Deep Learning ◽  
Network Model ◽  
Endoscopic Ultrasonography ◽  
Data Augmentation ◽  
Clinical Information ◽  
Training Data ◽  
Fine Tuning ◽  
Cystic Neoplasm ◽  
Cystic Neoplasms

Mucinous cystic neoplasms (MCN) and serous cystic neoplasms (SCN) account for a large portion of solitary pancreatic cystic neoplasms (PCN). In this study we implemented a convolutional neural network (CNN) model using ResNet50 to differentiate between MCN and SCN. The training data were collected retrospectively from 59 MCN and 49 SCN patients from two different hospitals. Data augmentation was used to enhance the size and quality of training datasets. Fine-tuning training approaches were utilized by adopting the pre-trained model from transfer learning while training selected layers. Testing of the network was conducted by varying the endoscopic ultrasonography (EUS) image sizes and positions to evaluate the network performance for differentiation. The proposed network model achieved up to 82.75% accuracy and a 0.88 (95% CI: 0.817–0.930) area under curve (AUC) score. The performance of the implemented deep learning networks in decision-making using only EUS images is comparable to that of traditional manual decision-making using EUS images along with supporting clinical information. Gradient-weighted class activation mapping (Grad-CAM) confirmed that the network model learned the features from the cyst region accurately. This study proves the feasibility of diagnosing MCN and SCN using a deep learning network model. Further improvement using more datasets is needed.

scholarly journals Decision support tools to improve cancer diagnostic decision making in primary care: a systematic review

2019 ◽  
Vol 69 (689) ◽  
pp. e809-e818 ◽  
Author(s):  
Sophie Chima ◽  
Jeanette C Reece ◽  
Kristi Milley ◽  
Shakira Milton ◽  
Jennifer G McIntosh ◽  
...  
Keyword(s):  
Systematic Review ◽  
Primary Care ◽  
Decision Making ◽  
Decision Support ◽  
Cost Effectiveness ◽  
Cancer Diagnosis ◽  
Decision Support Tools ◽  
Diagnostic Decision ◽  
Support Tools ◽  
Diagnostic Decision Making

BackgroundThe diagnosis of cancer in primary care is complex and challenging. Electronic clinical decision support tools (eCDSTs) have been proposed as an approach to improve GP decision making, but no systematic review has examined their role in cancer diagnosis.AimTo investigate whether eCDSTs improve diagnostic decision making for cancer in primary care and to determine which elements influence successful implementation.Design and settingA systematic review of relevant studies conducted worldwide and published in English between 1 January 1998 and 31 December 2018.MethodPreferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines were followed. MEDLINE, EMBASE, and the Cochrane Central Register of Controlled Trials were searched, and a consultation of reference lists and citation tracking was carried out. Exclusion criteria included the absence of eCDSTs used in asymptomatic populations, and studies that did not involve support delivered to the GP. The most relevant Joanna Briggs Institute Critical Appraisal Checklists were applied according to study design of the included paper.ResultsOf the nine studies included, three showed improvements in decision making for cancer diagnosis, three demonstrated positive effects on secondary clinical or health service outcomes such as prescribing, quality of referrals, or cost-effectiveness, and one study found a reduction in time to cancer diagnosis. Barriers to implementation included trust, the compatibility of eCDST recommendations with the GP’s role as a gatekeeper, and impact on workflow.ConclusioneCDSTs have the capacity to improve decision making for a cancer diagnosis, but the optimal mode of delivery remains unclear. Although such tools could assist GPs in the future, further well-designed trials of all eCDSTs are needed to determine their cost-effectiveness and the most appropriate implementation methods.

scholarly journals CrossGR

2021 ◽  
Vol 5 (1) ◽  
pp. 1-23
Author(s):  
Xinyi Li ◽  
Liqiong Chang ◽  
Fangfang Song ◽  
Ju Wang ◽  
Xiaojiang Chen ◽  
...  
Keyword(s):  
Gesture Recognition ◽  
Low Cost ◽  
User Involvement ◽  
Recognition System ◽  
Training Data ◽  
Generative Adversarial Network ◽  
Adversarial Network ◽  
Target User ◽  
Order Of Magnitude ◽  
Training Examples

This paper focuses on a fundamental question in Wi-Fi-based gesture recognition: "Can we use the knowledge learned from some users to perform gesture recognition for others?". This problem is also known as cross-target recognition. It arises in many practical deployments of Wi-Fi-based gesture recognition where it is prohibitively expensive to collect training data from every single user. We present CrossGR, a low-cost cross-target gesture recognition system. As a departure from existing approaches, CrossGR does not require prior knowledge (such as who is currently performing a gesture) of the target user. Instead, CrossGR employs a deep neural network to extract user-agnostic but gesture-related Wi-Fi signal characteristics to perform gesture recognition. To provide sufficient training data to build an effective deep learning model, CrossGR employs a generative adversarial network to automatically generate many synthetic training data from a small set of real-world examples collected from a small number of users. Such a strategy allows CrossGR to minimize the user involvement and the associated cost in collecting training examples for building an accurate gesture recognition system. We evaluate CrossGR by applying it to perform gesture recognition across 10 users and 15 gestures. Experimental results show that CrossGR achieves an accuracy of over 82.6% (up to 99.75%). We demonstrate that CrossGR delivers comparable recognition accuracy, but uses an order of magnitude less training samples collected from the end-users when compared to state-of-the-art recognition systems.

TMD Diagnostic Decision-Making and Probability Theory. Part II

CRANIO® ◽  
1996 ◽  
Vol 14 (4) ◽  
pp. 312-319
Author(s):  
L.V. Christensen ◽  
D.C. McKay
Keyword(s):  
Decision Making ◽  
Probability Theory ◽  
Diagnostic Decision ◽  
Diagnostic Decision Making

scholarly journals Generalization-Based Acquisition of Training Data for Motor Primitive Learning by Neural Networks

2021 ◽  
Vol 11 (3) ◽  
pp. 1013
Author(s):  
Zvezdan Lončarević ◽  
Rok Pahič ◽  
Aleš Ude ◽  
Andrej Gams
Keyword(s):  
Neural Networks ◽  
Dimensionality Reduction ◽  
Gaussian Process Regression ◽  
Search Space ◽  
Robot Learning ◽  
Training Data ◽  
Practical Applications ◽  
Latent Space ◽  
Real Robot ◽  
Low Dimensional

Autonomous robot learning in unstructured environments often faces the problem that the dimensionality of the search space is too large for practical applications. Dimensionality reduction techniques have been developed to address this problem and describe motor skills in low-dimensional latent spaces. Most of these techniques require the availability of a sufficiently large database of example task executions to compute the latent space. However, the generation of many example task executions on a real robot is tedious, and prone to errors and equipment failures. The main result of this paper is a new approach for efficient database gathering by performing a small number of task executions with a real robot and applying statistical generalization, e.g., Gaussian process regression, to generate more data. We have shown in our experiments that the data generated this way can be used for dimensionality reduction with autoencoder neural networks. The resulting latent spaces can be exploited to implement robot learning more efficiently. The proposed approach has been evaluated on the problem of robotic throwing at a target. Simulation and real-world results with a humanoid robot TALOS are provided. They confirm the effectiveness of generalization-based database acquisition and the efficiency of learning in a low-dimensional latent space.

Abstract 1122‐000047: Machine Learning to Predict Stroke Outcomes after Mechanical Thrombectomy

2021 ◽  
Vol 1 (S1) ◽  
Author(s):  
Mehdi Bouslama ◽  
Leonardo Pisani ◽  
Diogo Haussen ◽  
Raul Nogueira
Keyword(s):  
Machine Learning ◽  
Decision Making ◽  
Cerebral Artery ◽  
Training Data ◽  
Gradient Boosting ◽  
Support Vector ◽  
Anterior Circulation ◽  
Discriminative Performance ◽  
Post Procedure ◽  
Procedural Models

Introduction : Prognostication is an integral part of clinical decision‐making in stroke care. Machine learning (ML) methods have gained increasing popularity in the medical field due to their flexibility and high performance. Using a large comprehensive stroke center registry, we sought to apply various ML techniques for 90‐day stroke outcome predictions after thrombectomy. Methods : We used individual patient data from our prospectively collected thrombectomy database between 09/2010 and 03/2020. Patients with anterior circulation strokes (Internal Carotid Artery, Middle Cerebral Artery M1, M2, or M3 segments and Anterior Cerebral Artery) and complete records were included. Our primary outcome was 90‐day functional independence (defined as modified Rankin Scale score 0–2). Pre‐ and post‐procedure models were developed. Four known ML algorithms (support vector machine, random forest, gradient boosting, and artificial neural network) were implemented using a 70/30 training‐test data split and 10‐fold cross‐validation on the training data for model calibration. Discriminative performance was evaluated using the area under the receiver operator characteristics curve (AUC) metric. Results : Among 1248 patients with anterior circulation large vessel occlusion stroke undergoing thrombectomy during the study period, 1020 had complete records and were included in the analysis. In the training data (n = 714), 49.3% of the patients achieved independence at 90‐days. Fifteen baseline clinical, laboratory and neuroimaging features were used to develop the pre‐procedural models, with four additional parameters included in the post‐procedure models. For the preprocedural models, the highest AUC was 0.797 (95%CI [0.75‐ 0.85]) for the gradient boosting model. Similarly, the same ML technique performed best on post‐procedural data and had an improved discriminative performance compared to the pre‐procedure model with an AUC of 0.82 (95%CI [0.77‐ 0.87]). Conclusions : Our pre‐and post‐procedural models reliably estimated outcomes in stroke patients undergoing thrombectomy. They represent a step forward in creating simple and efficient prognostication tools to aid treatment decision‐making. A web‐based platform and related mobile app are underway.

Sign in / Sign up

Export Citation Format

Share Document