scholarly journals Deterministic Local Interpretable Model-Agnostic Explanations for Stable Explainability

2021 ◽  
Vol 3 (3) ◽  
pp. 525-541
Author(s):  
Muhammad Rehman Zafar ◽  
Naimul Khan
Keyword(s):  
Random Perturbation ◽  
Simulated Data ◽  
Training Data ◽  
Agglomerative Hierarchical Clustering ◽  
Local Algorithms ◽  
Feature Importance ◽  
Critical Issues ◽  
Interpretable Model ◽  
Synthetic Datasets ◽  
The Stability

Local Interpretable Model-Agnostic Explanations (LIME) is a popular technique used to increase the interpretability and explainability of black box Machine Learning (ML) algorithms. LIME typically creates an explanation for a single prediction by any ML model by learning a simpler interpretable model (e.g., linear classifier) around the prediction through generating simulated data around the instance by random perturbation, and obtaining feature importance through applying some form of feature selection. While LIME and similar local algorithms have gained popularity due to their simplicity, the random perturbation methods result in shifts in data and instability in the generated explanations, where for the same prediction, different explanations can be generated. These are critical issues that can prevent deployment of LIME in sensitive domains. We propose a deterministic version of LIME. Instead of random perturbation, we utilize Agglomerative Hierarchical Clustering (AHC) to group the training data together and K-Nearest Neighbour (KNN) to select the relevant cluster of the new instance that is being explained. After finding the relevant cluster, a simple model (i.e., linear model or decision tree) is trained over the selected cluster to generate the explanations. Experimental results on six public (three binary and three multi-class) and six synthetic datasets show the superiority for Deterministic Local Interpretable Model-Agnostic Explanations (DLIME), where we quantitatively determine the stability and faithfulness of DLIME compared to LIME.

Scalable Approach to High Coverages on Oxides via Iterative Training of a Machine-Learning Algorithm

2019 ◽  
Author(s):  
Andrew Medford ◽  
Shengchun Yang ◽  
Fuzhu Liu
Keyword(s):  
Machine Learning ◽  
Chemical Potential ◽  
Learning Algorithm ◽  
Absolute Error ◽  
Low Energy ◽  
Training Data ◽  
High Coverage ◽  
Metal Compounds ◽  
Adsorption Energies ◽  
The Stability

Understanding the interaction of multiple types of adsorbate molecules on solid surfaces is crucial to establishing the stability of catalysts under various chemical environments. Computational studies on the high coverage and mixed coverages of reaction intermediates are still challenging, especially for transition-metal compounds. In this work, we present a framework to predict differential adsorption energies and identify low-energy structures under high- and mixed-adsorbate coverages on oxide materials. The approach uses Gaussian process machine-learning models with quantified uncertainty in conjunction with an iterative training algorithm to actively identify the training set. The framework is demonstrated for the mixed adsorption of CH<sub>x</sub>, NH<sub>x</sub> and OH<sub>x</sub> species on the oxygen vacancy and pristine rutile TiO<sub>2</sub>(110) surface sites. The results indicate that the proposed algorithm is highly efficient at identifying the most valuable training data, and is able to predict differential adsorption energies with a mean absolute error of ~0.3 eV based on <25% of the total DFT data. The algorithm is also used to identify 76% of the low-energy structures based on <30% of the total DFT data, enabling construction of surface phase diagrams that account for high and mixed coverage as a function of the chemical potential of C, H, O, and N. Furthermore, the computational scaling indicates the algorithm scales nearly linearly (N<sup>1.12</sup>) as the number of adsorbates increases. This framework can be directly extended to metals, metal oxides, and other materials, providing a practical route toward the investigation of the behavior of catalysts under high-coverage conditions.

scholarly journals Holdout-Based Empirical Assessment of Mixed-Type Synthetic Data

2021 ◽  
Vol 4 ◽  
Author(s):  
Michael Platzer ◽  
Thomas Reutterer
Keyword(s):  
Mixed Type ◽  
Synthetic Data ◽  
Training Data ◽  
Privacy Risk ◽  
Individual Level ◽  
Empirical Assessment ◽  
Model Free ◽  
Private Data ◽  
Synthetic Datasets ◽  
The Individual

AI-based data synthesis has seen rapid progress over the last several years and is increasingly recognized for its promise to enable privacy-respecting high-fidelity data sharing. This is reflected by the growing availability of both commercial and open-sourced software solutions for synthesizing private data. However, despite these recent advances, adequately evaluating the quality of generated synthetic datasets is still an open challenge. We aim to close this gap and introduce a novel holdout-based empirical assessment framework for quantifying the fidelity as well as the privacy risk of synthetic data solutions for mixed-type tabular data. Measuring fidelity is based on statistical distances of lower-dimensional marginal distributions, which provide a model-free and easy-to-communicate empirical metric for the representativeness of a synthetic dataset. Privacy risk is assessed by calculating the individual-level distances to closest record with respect to the training data. By showing that the synthetic samples are just as close to the training as to the holdout data, we yield strong evidence that the synthesizer indeed learned to generalize patterns and is independent of individual training records. We empirically demonstrate the presented framework for seven distinct synthetic data solutions across four mixed-type datasets and compare these then to traditional data perturbation techniques. Both a Python-based implementation of the proposed metrics and the demonstration study setup is made available open-source. The results highlight the need to systematically assess the fidelity just as well as the privacy of these emerging class of synthetic data generators.

scholarly journals VEHICLE LOCALIZATION BY LIDAR POINT CORRELATION IMPROVED BY CHANGE DETECTION

2016 ◽  
Vol XLI-B1 ◽  
pp. 703-710 ◽  
Author(s):  
A. Schlichting ◽  
C. Brenner
Keyword(s):  
Change Detection ◽  
Reference Data ◽  
Correlation Method ◽  
Simulated Data ◽  
Detection Algorithm ◽  
Training Data ◽  
Image Correlation ◽  
Data Generation ◽  
Vehicle Localization ◽  
Dynamic Objects

LiDAR sensors are proven sensors for accurate vehicle localization. Instead of detecting and matching features in the LiDAR data, we want to use the entire information provided by the scanners. As dynamic objects, like cars, pedestrians or even construction sites could lead to wrong localization results, we use a change detection algorithm to detect these objects in the reference data. If an object occurs in a certain number of measurements at the same position, we mark it and every containing point as static. In the next step, we merge the data of the single measurement epochs to one reference dataset, whereby we only use static points. Further, we also use a classification algorithm to detect trees. <br><br> For the online localization of the vehicle, we use simulated data of a vertical aligned automotive LiDAR sensor. As we only want to use static objects in this case as well, we use a random forest classifier to detect dynamic scan points online. Since the automotive data is derived from the LiDAR Mobile Mapping System, we are able to use the labelled objects from the reference data generation step to create the training data and further to detect dynamic objects online. The localization then can be done by a point to image correlation method using only static objects. We achieved a localization standard deviation of about 5 cm (position) and 0.06° (heading), and were able to successfully localize the vehicle in about 93 % of the cases along a trajectory of 13 km in Hannover, Germany.

scholarly journals Generating stable molecules using imitation and reinforcement learning

2021 ◽  
Author(s):  
Søren Ager Meldgaard ◽  
Jonas Köhler ◽  
Henrik Lund Mortensen ◽  
Mads-Peter Verner Christiansen ◽  
Frank Noé ◽  
...  
Keyword(s):  
Reinforcement Learning ◽  
Chemical Space ◽  
Training Data ◽  
Graph Representation ◽  
Imitation Learning ◽  
Training Set ◽  
Machine Learning Methods ◽  
Multiple Copies ◽  
The Stability ◽  
3D Information

Abstract Chemical space is routinely explored by machine learning methods to discover interesting molecules, before time-consuming experimental synthesizing is attempted. However, these methods often rely on a graph representation, ignoring 3D information necessary for determining the stability of the molecules. We propose a reinforcement learning approach for generating molecules in cartesian coordinates allowing for quantum chemical prediction of the stability. To improve sample-efficiency we learn basic chemical rules from imitation learning on the GDB-11 database to create an initial model applicable for all stoichiometries. We then deploy multiple copies of the model conditioned on a specific stoichiometry in a reinforcement learning setting. The models correctly identify low energy molecules in the database and produce novel isomers not found in the training set. Finally, we apply the model to larger molecules to show how reinforcement learning further refines the imitation learning model in domains far from the training data.

scholarly journals On the utility of dreaming: A general model for how learning in artificial agents can benefit from data hallucination

2020 ◽  
pp. 105971231989648 ◽  
Author(s):  
David Windridge ◽  
Henrik Svensson ◽  
Serge Thill
Keyword(s):  
Machine Learning ◽  
Simulated Data ◽  
Training Data ◽  
Successful Implementation ◽  
Artificial Agents ◽  
Learning Context ◽  
Training Set ◽  
Convergence Point ◽  
And Training ◽  
General Method

We consider the benefits of dream mechanisms – that is, the ability to simulate new experiences based on past ones – in a machine learning context. Specifically, we are interested in learning for artificial agents that act in the world, and operationalize “dreaming” as a mechanism by which such an agent can use its own model of the learning environment to generate new hypotheses and training data. We first show that it is not necessarily a given that such a data-hallucination process is useful, since it can easily lead to a training set dominated by spurious imagined data until an ill-defined convergence point is reached. We then analyse a notably successful implementation of a machine learning-based dreaming mechanism by Ha and Schmidhuber (Ha, D., & Schmidhuber, J. (2018). World models. arXiv e-prints, arXiv:1803.10122). On that basis, we then develop a general framework by which an agent can generate simulated data to learn from in a manner that is beneficial to the agent. This, we argue, then forms a general method for an operationalized dream-like mechanism. We finish by demonstrating the general conditions under which such mechanisms can be useful in machine learning, wherein the implicit simulator inference and extrapolation involved in dreaming act without reinforcing inference error even when inference is incomplete.

Global Configuration Design of Umbilicals in Deepwater

2005 ◽  
Author(s):  
Ray Fleming ◽  
Thanos Moros ◽  
Rupak Ghosh ◽  
Kostas Lambrakos ◽  
Dave Robson
Keyword(s):  
Configuration Design ◽  
Loop Current ◽  
Global Configuration ◽  
Critical Issues ◽  
The Stability ◽  
Heave Motion ◽  
Vessel Motion ◽  
And Control ◽  
Selection Of

Global configuration design of subsea umbilical risers in deep water is a major challenge due to extreme environmental and operational requirements. The critical issues considered in design are the interference between umbilicals in the presence of strong loop and submerged current, and the on-bottom stability along with the strength and fatigue requirements. The vessel motion primarily controls the selection of the configuration, catenary or lazy wave, and the latter is an obvious choice in the presence of significant heave motion. The length and routing on the seabed ensures the on-bottom umbilical stability by dissipating the axial load through soil friction. However, the interference with other subsea components as well as the space availability can also be a governing criterion in the routing. Finally, all these design requirements must be satisfied economically from the perspective of overall cost of the project without compromising quality of the product and safety of design. This paper presents a brief outline of the global configuration design of umbilicals accounting for various design considerations. The host is a semi submersible in a water depth of 6050-ft in the Gulf of Mexico. The lazy wave configurations of the chemical injection and control umbilicals are considered for the study presented herein. The methodology of design for the global configuration is discussed considering different environmental loadings such as the 100-yr and 10-yr loop current, and 100-yr submerged current. The stability of the umbilical on the seabed is discussed on the basis of analysis results for the environmental loadings with dominant vessel motions. The phenomenon of “walking” under the influence of dynamic loading is investigated and the necessary considerations in design to prevent the umbilicals from “walking” are also discussed.

scholarly journals Kernel-Based Ensemble Learning in Python

Information ◽  
2020 ◽  
Vol 11 (2) ◽  
pp. 63 ◽  
Author(s):  
Benjamin Guedj ◽  
Bhargav Srinivasa Desikan
Keyword(s):  
Numerical Experiments ◽  
Learning Strategy ◽  
Learning Algorithm ◽  
Real Life ◽  
Training Data ◽  
Smoothing Kernel ◽  
Regression Problems ◽  
Classification And Regression ◽  
Synthetic Datasets ◽  
Python Package

We propose a new supervised learning algorithm for classification and regression problems where two or more preliminary predictors are available. We introduce KernelCobra, a non-linear learning strategy for combining an arbitrary number of initial predictors. KernelCobra builds on the COBRA algorithm introduced by Biau et al. (2016), which combined estimators based on a notion of proximity of predictions on the training data. While the COBRA algorithm used a binary threshold to declare which training data were close and to be used, we generalise this idea by using a kernel to better encapsulate the proximity information. Such a smoothing kernel provides more representative weights to each of the training points which are used to build the aggregate and final predictor, and KernelCobra systematically outperforms the COBRA algorithm. While COBRA is intended for regression, KernelCobra deals with classification and regression. KernelCobra is included as part of the open source Python package Pycobra (0.2.4 and onward), introduced by Srinivasa Desikan (2018). Numerical experiments were undertaken to assess the performance (in terms of pure prediction and computational complexity) of KernelCobra on real-life and synthetic datasets.

scholarly journals Deep Interest-Shifting Network with Meta-Embeddings for Fresh Item Recommendation

Complexity ◽  
2020 ◽  
Vol 2020 ◽  
pp. 1-13
Author(s):  
Zhao Li ◽  
Haobo Wang ◽  
Donghui Ding ◽  
Shichang Hu ◽  
Zhen Zhang ◽  
...  
Keyword(s):  
Real World ◽  
Contextual Information ◽  
Cold Start ◽  
Training Data ◽  
Learning To Learn ◽  
User Interests ◽  
Synthetic Datasets ◽  
Fresh Products ◽  
Embedding Performance ◽  
Deep Interest

Nowadays, people have an increasing interest in fresh products such as new shoes and cosmetics. To this end, an E-commerce platform Taobao launched a fresh-item hub page on the recommender system, with which customers can freely and exclusively explore and purchase fresh items, namely, the New Tendency page. In this work, we make a first attempt to tackle the fresh-item recommendation task with two major challenges. First, a fresh-item recommendation scenario usually faces the challenge that the training data are highly deficient due to low page views. In this paper, we propose a deep interest-shifting network (DisNet), which transfers knowledge from a huge number of auxiliary data and then shifts user interests with contextual information. Furthermore, three interpretable interest-shifting operators are introduced. Second, since the items are fresh, many of them have never been exposed to users, leading to a severe cold-start problem. Though this problem can be alleviated by knowledge transfer, we further babysit these fully cold-start items by a relational meta-Id-embedding generator (RM-IdEG). Specifically, it trains the item id embeddings in a learning-to-learn manner and integrates relational information for better embedding performance. We conducted comprehensive experiments on both synthetic datasets as well as a real-world dataset. Both DisNet and RM-IdEG significantly outperform state-of-the-art approaches, respectively. Empirical results clearly verify the effectiveness of the proposed techniques, which are arguably promising and scalable in real-world applications.

IMPROVING STABILITY OF DECISION TREES

2002 ◽  
Vol 16 (02) ◽  
pp. 145-159 ◽  
Author(s):  
MARK LAST ◽  
ODED MAIMON ◽  
EINAT MINKOV
Keyword(s):  
Decision Tree ◽  
Predictive Accuracy ◽  
Statistical Significance ◽  
Training Data ◽  
Tree Level ◽  
Information Theoretic ◽  
Tree Algorithms ◽  
Tree Models ◽  
The Stability ◽  
Information Theoretic Method

Decision-tree algorithms are known to be unstable: small variations in the training set can result in different trees and different predictions for the same validation examples. Both accuracy and stability can be improved by learning multiple models from bootstrap samples of training data, but the "meta-learner" approach makes the extracted knowledge hardly interpretable. In the following paper, we present the Info-Fuzzy Network (IFN), a novel information-theoretic method for building stable and comprehensible decision-tree models. The stability of the IFN algorithm is ensured by restricting the tree structure to using the same feature for all nodes of the same tree level and by the built-in statistical significance tests. The IFN method is shown empirically to produce more compact and stable models than the "meta-learner" techniques, while preserving a reasonable level of predictive accuracy.

RECOGNITION OF CONTACT STATE BY USING NEURAL NETWORK FOR MICROMACHINED ARRAY TYPE TACTILE SENSOR

2005 ◽  
Vol 02 (03) ◽  
pp. 181-190 ◽  
Author(s):  
SEIJI AOYAGI ◽  
TAKAAKI TANAKA ◽  
KENJI MAKIHIRA
Keyword(s):  
Neural Network ◽  
Rotation Angle ◽  
Simulated Data ◽  
Tactile Sensor ◽  
Training Data ◽  
Sensing Element ◽  
Array Type ◽  
Tactile Pattern ◽  
Hidden Layer ◽  
Contact Position

In this paper, a force sensing element having a pillar and a diaphragm is proposed and thereafter fabricated by micromachining. Piezo resistors are fabricated on a silicon diaphragm for detecting distortions caused by a force input to a pillar on the diaphragm. Since a practical arrayed sensor consisting of many of this element is still under development, the output of an assumed arrayed type tactile sensor is simulated by FEM (finite element method). Using simulated data, the possibility of tactile pattern recognition using a neural network (NN) is investigated. The learning method of NN, the number of units of the input layer and the hidden layer, as well as the number of training data are investigated for realizing high probability of recognition. The 14 subjects having different shape and size are recognized. This recognition succeeded even if the contact position and the rotation angle of these objects are changed.

Sign in / Sign up

Export Citation Format

Share Document