Filters








25,875 Hits in 4.1 sec

Visualizing the decision-making process in deep neural decision forest [article]

Shichao Li, Kwang-Ting Cheng
2019 arXiv   pre-print
In this work, we first trace the decision-making process of this model and visualize saliency maps to understand which portion of the input influence it more for both classification and regression problems  ...  Deep neural decision forest (NDF) achieved remarkable performance on various vision tasks via combining decision tree and deep representation learning.  ...  We gratefully acknowledge the support of NVIDIA Corporation with the donation of the Titan Xp GPU used for this research.  ... 
arXiv:1904.09201v1 fatcat:4zm5va2z2ng5vluuduw2nnj7ni

Facial age estimation by deep residual decision making [article]

Shichao Li, Kwang-Ting Cheng
2019 arXiv   pre-print
However, it has not been applied to deep neural decision forest (NDF).  ...  We further employ gradient-based technique to visualize the decision-making process of NDF and understand how it is influenced by facial image inputs.  ...  Methodology Residual Neural Decision Forest A deep neural decision forest (NDF) is an ensemble of deep neural decision trees. Each tree consists of splitting nodes and leaf nodes.  ... 
arXiv:1908.10737v1 fatcat:o44hvmdfivgu5khpvklrxjz3ie

DS_10.1177_0022034520969115 – Supplemental material for Application of Artificial Intelligence in Dentistry

T. Shan, F.R. Tay, L. Gu
2020 Figshare  
Supplemental material, DS_10.1177_0022034520969115 for Application of Artificial Intelligence in Dentistry by T. Shan, F.R. Tay and L. Gu in Journal of Dental Research  ...  By building a forest, the mistakes of a single decision tree are compensated for by other decision trees in the forest (https://www.sciencedirect.com/science/article/pii/S1566253515000561; accessed 6-2  ...  A decision forest consists of several decision trees in which their predictions are combined into a final prediction.  ... 
doi:10.25384/sage.13168817.v1 fatcat:wivudjmwpbhi3drui6r3hhoekq

A copula-based visualization technique for a neural network [article]

Yusuke Kubo, Yuto Komori, Toyonobu Okuyama, Hiroshi Tokieda
2020 arXiv   pre-print
However, a neural network is not considered interpretable due to the ambiguity in its decision-making process.  ...  Therefore, in this study, we propose a new algorithm that reveals which feature values the trained neural network considers important and which paths are mainly traced in the process of decision-making  ...  CVT can easily visualize the decision-making process of a neural network using correlation coefficients.  ... 
arXiv:2003.12317v1 fatcat:u7gbjmv4z5ggzop3nc5yuvcp5m

Enhanced Multi-dimensional and Multi-grained Cascade Forest for Cloud/snow Recognition Using Multispectral Satellite Remote Sensing Imagery

Meng Xia, Zhijie Wang, Fang Han, Yangting Kang
2021 IEEE Access  
In addition, based on the tree-based structure, the proposed method well balances the performance and efficiency of cloud/snow recognition, which can be considered as an alternative to the Neural Network  ...  The multi-dimensional deep forest structure with the representation learning ability allows it to capture the spatial and spectral information of cloud/snow satellite imagery accordingly equipped with  ...  There is a big difference between Cascade Forest and the Deep Neural Network. The feature extraction process of the Deep Neural Network is guided and updated by highlevel error backpropagation.  ... 
doi:10.1109/access.2021.3114185 fatcat:bye2bysql5eipdhd2nlxblszvu

Learning Representations for Axis-Aligned Decision Forests through Input Perturbation [article]

Sebastian Bruch, Jan Pfeifer, Mathieu Guillame-bert
2020 arXiv   pre-print
Our model is simply a decision forest, possibly trained using any forest learning algorithm, atop a deep neural network.  ...  Our framework has the advantage that it is applicable to any arbitrary decision forest and that it allows the use of arbitrary deep neural networks for representation learning.  ...  Evaluation with synthetic datasets allows us to design patterns that are difficult to model with a decision forest alone, and makes it possible to visually inspect the output of the model.  ... 
arXiv:2007.14761v2 fatcat:ohiayzpetfelvbfocm6aquxfpq

Interpreting Deep Learning Model Using Rule-based Method [article]

Xiaojian Wang, Jingyuan Wang, Ke Tang
2020 arXiv   pre-print
In this paper, we propose a multi-level decision framework to provide comprehensive interpretation for the deep neural network model.  ...  In the evaluation process, both functionally-grounded and human-grounded methods are used to ensure credibility.  ...  In terms of global explanation, MLD framework aims at finding the important features in the neural network decision process.  ... 
arXiv:2010.07824v1 fatcat:plkdyopbyvhlxgjdkae3hmolky

GrCAN: Gradient Boost Convolutional Autoencoder with Neural Decision Forest [article]

Manqing Dong, Lina Yao, Xianzhi Wang, Boualem Benatallah, Shuai Zhang
2018 arXiv   pre-print
While the random forest is robust irrespective of the data domain, the deep neural network has advantages in handling high dimensional data.  ...  Random forest and deep neural network are two schools of effective classification methods in machine learning.  ...  [4] : they proposed deep neural decision forests, which is also used in our model.  ... 
arXiv:1806.08079v2 fatcat:vvp7v3utdnce7jnezbfxtvh6re

A Survey and Comparison of Artificial Intelligence Techniques for Image Classification and Their Applications

2016 International Journal of Science and Research (IJSR)  
This paper focuses on a case study of a number of image classification algorithms which include decision trees, k-nearest neighbours, deep neural networks, Convolutional neural networks, Support vector  ...  machines and random forest.  ...  process of pretraining deep algorithms. [18] Convolutional Neural networks Convolutional neural networks are best suited for image recognition, as they retain the spatial topology [19] .  ... 
doi:10.21275/v5i4.nov162497 fatcat:j7gylplrvbexritnb7pmte24je

Analysis and Prediction of Graduate Admissions Based on Pre-COVID and post-COVID Scenario

2021 International Journal of Advanced Trends in Computer Science and Engineering  
Various models such as Logistic Regression, Decision Tree, Random Forest, Gaussian Naive Bayes and Artificial Neural Networks are used to determine the change in probability of admission due to the effect  ...  on the chance of admit and can take preemptive decisions to facilitate the process.  ...  helps aid the admissions committee make the process of reviewing admission applications more efficient.  ... 
doi:10.30534/ijatcse/2021/051062021 fatcat:j3hnncj6lbhanbsdok4ri4glwm

Stop ordering machine learning algorithms by their explainability! A user-centered investigation of performance and explainability

Lukas-Valentin Herm, Kai Heinrich, Jonas Wanner, Christian Janiesch
2022 International Journal of Information Management  
Machine learning algorithms enable advanced decision making in contemporary intelligent systems. Research indicates that there is a tradeoff between their model performance and explainability.  ...  Second, we address the problem of end user perceptions of explainable artificial intelligence augmentations aimed at increasing the understanding of the decision logic of high-performing complex models  ...  Acknowledgement This research and development project is funded by the Bayerische Staatsministerium für Wirtschaft, Landesentwicklung und Energie (StMWi) within the framework concept "Informations-und  ... 
doi:10.1016/j.ijinfomgt.2022.102538 fatcat:m4niks5k6vehtja4ntkc4jarbu

Faultiness Decision Making for False Information in Online: A Systematic Approach

Yasir Babiker Hamdan, Sathish
2021 Journal of Soft Computing Paradigm  
The deep decision making section compares the input and make the decision wisely and it provides the more accurate output rather than single classifiers in deep learning.  ...  The two different approaches detect fake news in online and it gives to decision making section which is designed at tail in our research.  ...  Neural Network classification c) Decision making Section The decision making section accomplish deep learning from positive and negative reinforcement after combining the two effective algorithm to  ... 
doi:10.36548/jscp.2020.4.004 fatcat:uxe3xlhvi5hgbpue6yutmxzpzq

Applications of Explainable Artificial Intelligence in Diagnosis and Surgery

Yiming Zhang, Ying Weng, Jonathan Lund
2022 Diagnostics  
Compared with AI techniques such as deep learning, XAI can provide both decision-making and explanations of the model.  ...  In recent years, artificial intelligence (AI) has shown great promise in medicine. However, explainability issues make AI applications in clinical usages difficult.  ...  Conflicts of Interest: The authors declare no conflict of interest.  ... 
doi:10.3390/diagnostics12020237 pmid:35204328 pmcid:PMC8870992 fatcat:fk5gbai6szf2vhf222o7p6nkqy

Opening the Black Box of Financial AI with CLEAR-Trade: A CLass-Enhanced Attentive Response Approach for Explaining and Visualizing Deep Learning-Driven Stock Market Prediction

Devinder Kumar, Graham W. Taylor, Alexander Wong
2017 Journal of Computational Vision and Imaging Systems  
The resultsdemonstrate that CLEAR-Trade can provide significant insightinto the decision-making process of deep learning-driven financialmodels, particularly for regulatory processes, thus improving theirpotential  ...  In particular, CLEAR-Trade provides a effectiveway to visualize and explain decisions made by deep stock marketprediction models.  ...  The proposed CLEAR-Trade visualization framework improves financial model interpretability by providing effective visual interpretations of the decision-making process.  ... 
doi:10.15353/vsnl.v3i1.166 fatcat:keuf2jomhbby3e4satb3l67jpi

Applied Machine Learning and Artificial Intelligence in Rheumatology

Maria Hügle, Patrick Omoumi, Jaap van Laar, Joschka Boedecker, Thomas Hügle
2020 Rheumatology Advances in Practice  
Thus, in future shared decision-making will not only include the patient's opinion and the rheumatologist's empirical and evidence-based experience, but it will also be influenced by machine-learned evidence  ...  In the future, machine learning will likely assist rheumatologists in predicting the course of the disease and identifying important disease factors.  ...  The dynamic deep neural network architecture outperformed random forests and fully-connected neural networks, achieving a MSE of 0.9, which corresponds to an error of 8% in the range of the target value  ... 
doi:10.1093/rap/rkaa005 pmid:32296743 pmcid:PMC7151725 fatcat:76uspjabtfamfm42cmlsx4c5g4
« Previous Showing results 1 — 15 out of 25,875 results