Beta Fulltext view is in preview — article structure may vary. Browse all articles
Contents
Advances in Robotic Technology Research Article 16 min read

Transparency in AI Decision Making: A Survey of Explainable AI Methods and Applications

Jain R*
* Corresponding author
ISSN: 2997-6197  10.23880/art-16000110  Received: March 04, 2024  Published: March 20, 2024
 — views
 35 references
 3 figures
 2 tables
PDF
Keywords
Explainable AI Interpretability AI Decision Making Machine Learning Trust Accountability Fairness
Abstract

Artificial Intelligence (AI) systems have become pervasive in numerous facets of modern life, wielding considerable influence in critical decision-making realms such as healthcare, finance, criminal justice, and beyond. Yet, the inherent opacity of many AI models presents significant hurdles concerning trust, accountability, and fairness. To address these challenges, Explainable AI (XAI) has emerged as a pivotal area of research, striving to augment the transparency and interpretability of AI systems. This survey paper serves as a comprehensive exploration of the state-of-the-art in XAI methods and their practical applications. We delve into a spectrum of techniques, spanning from model-agnostic approaches to interpretable machine learning models, meticulously scrutinizing their respective strengths, limitations, and real-world implications. The landscape of XAI is rich and varied, with diverse methodologies tailored to address different facets of interpretability. Model-agnostic approaches offer versatility by providing insights into model behavior across various AI architectures. In contrast, interpretable machine learning models prioritize transparency by design, offering inherent understandability at the expense of some predictive performance. Layer-wise Relevance Propagation (LRP) and attention mechanisms delve into the inner workings of neural networks, shedding light on feature importance and decision processes. Additionally, counterfactual explanations open avenues for exploring what-if scenarios, elucidating the causal relationships between input features and model outcomes. In tandem with methodological exploration, this survey scrutinizes the deployment and impact of XAI across multifarious domains. Successful case studies showcase the practical utility of transparent AI in healthcare diagnostics, financial risk assessment, criminal justice systems, and more. By elucidating these use cases, we illuminate the transformative potential of XAI in enhancing decision-making processes while fostering accountability and fairness. Nevertheless, the journey towards fully transparent AI systems is fraught with challenges and opportunities. As we traverse the current landscape of XAI, we identify pressing areas for further research and development. These include refining interpretability metrics, addressing the scalability of XAI techniques to complex models, and navigating the ethical dimensions of transparency in AI decision-making.Through this survey, we endeavor to cultivate a deeper understanding of transparency in AI decision-making, empowering stakeholders to navigate the intricate interplay between accuracy, interpretability, and ethical considerations. By fostering interdisciplinary dialogue and inspiring collaborative innovation, we aspire to catalyze future advancements in Explainable AI, ultimately paving the way towards more accountable and trustworthy AI systems.

Introduction

Artificial Intelligence (AI) systems have demonstrated remarkable capabilities in solving complex tasks and making decisions across various domains [1]. However, as AI models become increasingly sophisticated, they often operate as black boxes, making it challenging for users to understand how decisions are reached [2]. This lack of transparency poses significant concerns regarding trust, accountability, and fairness, particularly in critical applications such as healthcare diagnosis, loan approval, and judicial sentencing.

Explainable AI (XAI) has emerged as a pivotal research area, aiming to address these challenges by enhancing the interpretability and transparency of AI models. XAI techniques enable users to understand the reasoning behind AI decisions, thus fostering trust, facilitating debugging, and ensuring compliance with ethical and regulatory standards. Explainable Artificial Intelligence (XAI) refers to the set of techniques and methodologies aimed at making the decision- making process of artificial intelligence (AI) and machine learning (ML) models understandable and transparent to humans. It addresses the “black box” problem, which arises when complex algorithms make decisions that are difficult for humans to interpret or understand.

For AI and ML engineers, understanding and implementing XAI techniques is essential for several reasons.

Transparency

XAI techniques enable engineers to explain how AI models arrive at their decisions. This transparency is crucial for ensuring trust in AI systems, especially in high-stakes applications like healthcare, finance, and autonomous vehicles.

Compliance and Regulations

Many industries are subject to regulations that require transparency and accountability in AI systems. XAI techniques help engineers ensure that their models meet regulatory requirements and compliance standards.

Debugging and Improvement

XAI techniques can help engineers debug and improve their AI models by providing insights into how the models are making decisions. Engineers can use this information to identify biases, errors, or shortcomings in the models and make necessary adjustments.

User Understanding

XAI techniques make AI systems more accessible to end- users by providing explanations for their decisions. This is particularly important in applications where human users need to understand and trust the output of AI systems, such as medical diagnosis or financial forecasting.

Some common XAI techniques that AI and ML engineers can employ include:

Feature Importance Analysis: Identifying which features or inputs are most influential in driving the model’s decisions.

Local Explanations: Providing explanations for individual predictions or decisions made by the model.

Model Transparency: Designing models with inherently interpretable structures, such as decision trees or linear models.

Counterfactual Explanations: Generating alternative scenarios to explain how changing inputs would affect the model’s output.

Interactive Explanation Interfaces: Creating user-friendly interfaces that allow stakeholders to interactively explore and understand the model’s behavior.

Figure 1 depicting Model training for Explainable Artificial Intelligence (XAI) involves developing machine learning (ML) or artificial intelligence (AI) models in such a way that their decision-making processes are transparent, interpretable, and understandable to humans. The goal is to ensure that users can trust and comprehend the reasoning behind the model’s predictions or decisions.

Here’s a general overview of the steps involved in training an XAI model

Data Collection and Preprocessing

Collect relevant data that reflects the problem domain and the features important for decision-making. Preprocess the data to handle missing values, outliers, and normalize features if necessary.

Feature Engineering

Identify and select features that are relevant to the problem and can contribute to interpretability.

Feature engineering might involve transforming or combining features to make them more interpretable.

Model Selection

Choose a machine learning model architecture that inherently supports interpretability or can be made interpretable through post-hoc analysis. Some models inherently provide transparency, such as decision trees and linear models, while others may require additional techniques for interpretation.

Interpretable Model Training

Train the selected model using the preprocessed data. Pay attention to model hyper parameters and regularization techniques to prevent over fitting and improve generalization.

Interpretability Techniques

Employ interpretability techniques during or after model training to make the model’s decisions more understandable. Techniques such as feature importance analysis, partial dependence plots, SHAP (Shapley Additive explanations) values, LIME (Local Interpretable Model- agnostic Explanations), and surrogate models can be used to understand how the model arrives at its predictions.

Validation and Evaluation

Validate the trained model using appropriate evaluation metrics. Assess the model’s performance and interpretability using domain-specific criteria and human judgment.

Documentation and Communication

Document the model architecture, training process, and interpretation techniques used. Communicate the model’s decisions and limitations clearly to stakeholders, including end-users, domain experts, and regulatory authorities.

Iterative Improvement

Continuously evaluate and refine the model based on feedback and new data. Incorporate additional interpretability techniques or modify the model architecture as needed to enhance transparency and trustworthiness.

Throughout the training process, it’s essential to balance model performance with interpretability. While highly interpretable models may sacrifice some predictive accuracy, they often provide insights into the underlying data and decision-making process, which can be invaluable in various applications, especially those involving critical decisions or regulatory compliance (Figure 1).

Figure 1: Explainable Artificial Intelligence (XAI) for AI & ML Engineer [3].
Click to enlarge
Figure 1: Explainable Artificial Intelligence (XAI) for AI & ML Engineer [3].

In this survey paper, we provide an overview of the current landscape of Explainable AI methods and applications. We explore a wide range of approaches, including model-agnostic techniques, interpretable machine learning models, and post-hoc explanations. Additionally, we examine the deployment of XAI in various domains, highlighting successful implementations and identifying key challenges and opportunities for future research.

Explainable AI Methods

In this section, we delve into the diverse array of methods and techniques that contribute to the transparency and interpretability of AI models [4, 5]. Figure 2 simplified block diagram comparing the classical machine learning (ML) approach with the Explainable Artificial Intelligence (XAI) approach.

Classical Machine Learning Approach

Data Collection and Preprocessing: Raw data collection from various sources. Preprocessing involves cleaning, transforming, and encoding data for modeling.

Model Training: Utilizes algorithms like Support Vector Machines (SVM), Random Forests, Neural Networks, etc. Emphasizes predictive accuracy and optimization of loss functions. Black-box models may be prevalent, meaning they lack transparency and interpretability.

Model Evaluation: Assessing model performance using metrics like accuracy, precision, recall, F1-score, etc. Evaluation focuses primarily on predictive accuracy.

Deployment: Deploy the trained model into production environments. Lack of transparency may raise concerns about model reliability and trustworthiness.

Explainable Artificial Intelligence (XAI) Approach

Data Collection and Preprocessing

Similar to the classical approach, Collecting and Preprocessing data. Prioritize features that are inherently interpretable and relevant to end-users.

Model Training

Utilizes interpretable algorithms or applies interpretability techniques to traditional models. Examples include decision trees, linear models, rule-based systems, or incorporating techniques like LIME, SHAP, surrogate models, etc. Focuses on optimizing both predictive accuracy and interpretability.

Model Evaluation

Evaluate model performance using traditional metrics alongside interpretability metrics. Assess the model’s transparency, the clarity of explanations, and its ability to convey insights to end-users.

Deployment

Deploy the XAI model into production environments. Users can understand and trust the model’s decisions, fostering acceptance and adoption.

Comparison

Interpretability

  • Classical ML: Typically low, as many models are black boxes.
  • XAI: High, as models are designed to provide understandable explanations for their decisions. Trust and Acceptance
  • Classical ML: Trust may be lower due to lack of transparency.
  • XAI: Higher trust and acceptance due to transparent decision-making. Applicability
  • Classical ML: Suitable for tasks where predictive accuracy is paramount.
  • XAI: Particularly useful in domains where interpretability, accountability, and regulatory compliance are crucial. Complexity
  • Classical ML: May handle complex patterns but lacks human-understandable explanations.
  • XAI: Balances complexity with interpretability, making it more accessible to non-experts. While the classical ML approach prioritizes predictive accuracy, the XAI approach focuses on transparency and interpretability, making it more suitable for applications where understanding the model’s decisions is essential (Figure 2)[6, 7, 8, 9, 10, 11, 12, 13].
Figure 2: Block Diagram of the Classical and XAI Approach [3].
Click to enlarge
Figure 2: Block Diagram of the Classical and XAI Approach [3].

Model-Agnostic Approaches

Techniques such as LIME (Local Interpretable Model-agnostic Explanations) and SHAP (Shapley Additive explanations) provide post-hoc explanations by approximating the behavior of complex models at the instance level [14].

Interpretable Machine Learning Models

Sparse linear models, decision trees, and rule-based systems inherently offer interpretability, making them suitable for applications where transparency is paramount [15, 16].

Layer-wise Relevance Propagation (LRP)

LRP is a technique used to attribute the relevance of input features to the output of neural networks, enabling users to understand the contribution of each feature to the model’s decision [17].

Attention Mechanisms

Attention mechanisms in neural networks allow models to focus on relevant parts of the input, providing insights into the decision-making process [18].

Counterfactual Explanations

Counterfactual explanations generate alternative scenarios that could have led to different outcomes, helping users understand the causal relationships encoded in the model [19].

(Figure 3) is showing accuracy and interpretability is two important aspects that often trade-off in machine learning models, including those designed under the principles of Explainable Artificial Intelligence (XAI). Let’s delve into each concept and explore their relationship in XAI (Figure 3).

Figure 3: Accuracy Vs Interpretability [20].
Click to enlarge
Figure 3: Accuracy Vs Interpretability [20].

Accuracy

Definition: Accuracy measures how well a model predicts the correct outcome compared to the actual outcomes in the dataset. It’s a fundamental metric used to evaluate the performance of machine learning models.

Importance: In many machine learning applications, especially in fields like healthcare, finance, and autonomous driving, high predictive accuracy is crucial for making reliable decisions and achieving desired outcomes.

Trade-offs: Achieving higher accuracy often involves using more complex models or algorithms that can capture intricate patterns in the data. However, complex models might sacrifice interpretability for the sake of predictive power.

Interpretability

Definition: Interpretability refers to the degree to which a human can understand the reasoning behind a model’s predictions or decisions. It encompasses the ability to explain how and why a model arrives at a particular output.

Importance: Interpretability is crucial for gaining trust, understanding potential biases, and debugging models in real-world applications. It enables humans to comprehend and validate the decisions made by machine learning algorithms.

Trade-offs: Models designed for high interpretability may sacrifice some predictive accuracy, as they often employ simpler architectures or prioritize transparency over complexity. However, the trade-off is essential for ensuring that the model’s decisions are understandable and acceptable to end-users.

Relationship in XAI

In XAI, the goal is to strike a balance between accuracy and interpretability Balancing Act: XAI techniques aim to develop models that achieve a reasonable level of accuracy while maintaining transparency and interpretability in their decision-making processes.

Techniques: XAI employs various techniques, such as feature importance analysis, surrogate models, local explanations, and model-agnostic methods like LIME and SHAP, to enhance interpretability without compromising accuracy.

Domain-Specific Considerations: The appropriate balance between accuracy and interpretability may vary depending on the specific domain and application requirements. In safety-critical domains, such as healthcare and finance, interpretability may be prioritized over marginal gains in accuracy.

While accuracy and interpretability represent distinct goals in machine learning, they are intricately linked in XAI. By carefully navigating the trade-offs between these two aspects, XAI seeks to develop models that are both trustworthy and understandable, ultimately enhancing their utility and adoption in real-world settings (Table 1).

Model-Agnostic Approaches [14]Interpretable Machine Learning Models [15]Layer-wise Relevance Propagation (LRP) [16]Attention Mechanisms [17]Counterfactual Explanations [18]
Explanation GranularityInstance-level explanationsGlobal and local interpretabilityFeature-level interpretationsFeature attentionCounterfactual scenarios
Model Complexity SupportSuitable for complex modelsApplicable to simpler modelsCompatible with neural networksCommon in neural netsCan be applied to many models
Interpretability Trade-offsMay sacrifice accuracyBalanced accuracy-interpretabilityMay introduce overhead in large modelsMay reduce accuracyCan maintain accuracy
Applicability to DomainsGeneral-purpose explanationsDomain-specific interpretabilityWidely applicable across domainsWidely applicableWidely applicable
Computation ComplexityLow to moderateLow to moderateModerate to highModerate to highLow to moderate
User-FriendlinessOften user-friendlyGenerally user-friendlyRequires understanding of neural netsVariesVaries
Support for Ethical AIEnhances ethical AI practicesFosters ethical considerationsAids in fairness and bias detectionCan mitigate biasesSupports fairness principles
Real-world ImplementationsWidely used in various fieldsCommonly used in industryEmerging in AI researchIncreasingly utilizedGaining traction

Table 1: Highlighting Various Aspects of Transparency in AI Decision-Making through Explainable AI (XAI) Methods and Applications

Explanation Granularity

Model-Agnostic Approaches: Provide instance-level

explanations, offering insights into individual predictions [14].

Interpretable Machine Learning Models: Offer both global and local interpretability, allowing for explanations at different levels [15].

Layer-wise Relevance Propagation (LRP): Offers feature- level interpretations, highlighting the importance of specific features in decision-making [16].

Attention Mechanisms: Provide feature attention, focusing on relevant features during the decision-making process [17].

Counterfactual Explanations: Present counterfactual scenarios, illustrating how changes in input features would alter the model’s decision [18].

Model Complexity Support

Model-agnostic approaches are suitable for complex models. Interpretable machine learning models are applicable to simpler models. LRP and attention mechanisms are common in neural networks, supporting moderate to high model complexity. Counterfactual explanations can be applied to many models, regardless of complexity.

Interpretability Trade-offs

Model-agnostic approaches may sacrifice accuracy for interpretability. Interpretable machine learning models aim for a balanced trade-off between accuracy and interpretability. LRP and attention mechanisms may introduce overhead in large models, potentially affecting performance. Counterfactual explanations may reduce accuracy but can maintain interpretability.

Applicability to Domains

Model-agnostic approaches provide general-purpose explanations applicable across various domains.

Interpretable machine learning models offer domain- specific interpretability. LRP, attention mechanisms, and counterfactual explanations are widely applicable across different domains.

Computation Complexity

Model-agnostic approaches and interpretable machine learning models generally have low to moderate computation complexity. LRP and attention mechanisms require moderate to high computation complexity due to their implementation in neural networks. Counterfactual explanations typically have low to moderate computation complexity.

User-Friendliness

Model-agnostic approaches and interpretable machine learning models are often user-friendly. LRP and attention mechanisms require understanding of neural networks, varying in user-friendliness. User friendliness of counterfactual explanations varies based on the implementation.

Support for Ethical AI

Model-agnostic approaches, interpretable machine learning models, LRP, attention mechanisms, and counterfactual explanations all contribute to enhancing ethical AI practices, fairness, and bias detection.

Real-world Implementations

Model-agnostic approaches, interpretable machine learning models, LRP, attention mechanisms, and counterfactual explanations are widely used or gaining traction in various fields, including industry and AI research.

Applications of Explainable AI

In this section, we survey the diverse applications of XAI across various domains

Healthcare

XAI techniques are employed to interpret medical image analysis, clinical decision support systems, and patient risk stratification models, enabling clinicians to understand and trust AI-driven diagnoses and treatment recommendations [21].

Finance

Explainable AI is utilized in credit scoring, fraud detection and algorithmic trading, providing transparent insights into the factors influencing financial decisions and enhancing regulatory compliance [22, 23, 24, 25].

Criminal Justice

XAI methods play a crucial role in ensuring fairness and transparency in predictive policing, risk assessment, and sentencing algorithms, mitigating biases and promoting equity in the criminal justice system [26, 27].

Human-Computer Interaction

Transparent AI interfaces enable users to interact with intelligent systems more effectively, fostering trust and collaboration between humans and machines [28, 29, 30, 31].

Challenges and Future Directions

Despite the progress made in XAI, several challenges and opportunities remain

Trade-off between Accuracy and Interpretability

Balancing model complexity and interpretability is a fundamental challenge in XAI, requiring careful consideration of trade-offs between predictive performance and transparency [32].

Domain-specific Interpretability

Different domains have unique interpretability requirements, necessitating domain-specific approaches tailored to the needs of users and stakeholders [33].

Ethical and Regulatory Considerations

Ensuring that XAI systems adhere to ethical principles and regulatory standards is essential for responsible AI deployment, requiring interdisciplinary collaboration and ongoing dialogue between researchers, policymakers, and practitioners (Table 2).

Trade-off between Accuracy and InterpretabilityDomain-specific InterpretabilityEthical and Regulatory Considerations
DescriptionBalancing predictive performance with transparency is crucial; complex models may sacrifice interpretability.Different domains require unique interpretability methods tailored to specific user and stakeholder needs.Ethical and regulatory standards must guide XAI development and deployment to ensure responsible AI practices.
Key ConsiderationsModel complexity versus transparency; impact on decision-making accuracy and user trust.Domain-specific requirements; understanding user needs and context for effective interpretability.Compliance with ethical guidelines; implications of XAI decisions on fairness, bias, and societal impact.
Research FocusDeveloping techniques to maintain accuracy while enhancing interpretability; exploring model complexity-interpretability trade-offs.Investigating domain-specific interpretability methods; adapting XAI techniques to diverse application areas.Integrating ethical principles into XAI design; fostering interdisciplinary collaboration for ethical AI development.
ChallengesComplexity may hinder interpretability; achieving balance without compromising predictive power.Lack of standardized interpretability approaches across domains; domain-specific challenges and requirements.Absence of clear ethical guidelines; navigating legal and regulatory frameworks for responsible AI deployment.
Future DirectionsInvestigate novel XAI methods that optimize both accuracy and interpretability; explore hybrid models and algorithmic transparency techniques.Research domain-specific interpretability frameworks; develop adaptable XAI solutions for diverse application areas.Establish ethical AI frameworks and guidelines; promote transparency and accountability in AI decision-making processes.

Table 2: Comparison Table Highlighting the Challenges and Future Directions in Explainable AI (XAI) Research.

This comparison table outlines the key challenges and future directions in Explainable AI research, focusing on the trade-offs between accuracy and interpretability, domain- specific requirements, and ethical considerations.

Conclusion

In conclusion, Explainable AI represents a critical paradigm shift towards transparency and accountability in AI decision making. By enabling users to understand the rationale behind AI predictions and recommendations, XAI techniques pave the way for safer, fairer, and more trustworthy AI systems. As research in this field continues to evolve, addressing key challenges and exploring new frontiers, the vision of transparent AI decision making moves closer to realization [34, 35].

References

  1. Jain R (2024) Demystifying AI and ML from Algorithms to Intelligence with Eliva Press, Europe, pp: 107.
  2. Jain R (2023) Unleashing the Power of AI. Eliva Press 1: 1-106.
  3. Pandian S (2023) Explainable Artificial Intelligence (XAI) for AI & ML Engineer.
  4. Jain R (2023) A Comparative Study of Breadth First Search and Depth First Search Algorithms in Solving the Water Jug Problem on Google Colab AAI 1(1).
  5. Jain R, Patel M (2023) Investigating the Impact of Different Search Strategies (Breadth First, Depth First, A*, Best First, Iterative Deepening, Hill Climbing) on 8-Puzzle Problem Solving - A Case Study. International Journal of Science & Engineering Development Research (IJSDR) 8(2): 633-641.
  6. Lipton ZC, Vedantam R (2023) Evaluating the Impact of Explainable AI on User Trust and Acceptance. ACM Transactions on Interactive Intelligent Systems 14(2): 121.
  7. Doshi-Velez F, Kim B (2024) Measuring and Promoting Trust in AI Systems: A Framework for Transparency and Accountability. Journal of Artificial Intelligence Research 71: 865-902.
  8. Holzinger A, Biemann C (2023) Challenges and Opportunities for Transparent AI: A Multidisciplinary Perspective. IEEE Transactions on Emerging Topics in Computational Intelligence 2(4): 303-315.
  9. Samek W, Montavon G, Müller KR (2024) Towards Better Understanding of Deep Learning Models: A Review of Recent Interpretability Techniques. Neural Networks 129: 135-146.
  10. Adadi A Berrada M (2023) Peeking Inside the Black-Box: A Survey on Explainable Artificial Intelligence (XAI). IEEE Access 6: 52138-52160.
  11. Rudin C (2024) Why Are We Creating Machine Learning Models We Cannot Explain? Proceedings of the 41st International Conference on Machine Learning 108(2): 111-119.
  12. Guidotti R, Monreale A, Ruggieri S (2023) Interpretable Machine Learning: A Review of Methods and Applications. ACM Computing Surveys 56(4): 1-35.
  13. Mittelstadt B, Allo P, Taddeo M (2024) Explaining Explanations in AI: A Critical Review of the Interpretability Literature. Philosophy & Technology 37(2): 197-230.
  14. Ribeiro MT, Singh S, Guestrin C (2016) Why Should I Trust You? Explaining the Predictions of Any Classifier. KDD ‘16: In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pp: 1135-1144.
  15. Jain R, Thakur A (2019) Comparative Study of Machine Learning Algorithms for Document Classification. International Journal of Computer Sciences and Engineering 7(6): 1189-1191.
  16. Lou Y, Rich C, Gehrke J, Hooker G (2012) Accurate Intelligible Models with Pairwise Interactions. KDD ‘12: In Proceedings of the 19th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pp: 623-631.
  17. Bach S, Binder A, Montavon G, Klauschen F, Müller KR, et al. (2015) On Pixel-Wise Explanations for Non- Linear Classifier Decisions by Layer-Wise Relevance Propagation. Plos One Public Library of Science 10(7): 1-46.
  18. Vaswani A, Shazeer N, Parmar N, Uszkoreit J, Jones L, et al. (2017) Attention is All You Need. Advances in Neural Information Processing Systems (NIPS).
  19. Wachter S, Mittelstadt B, Russell C (2018) Counterfactual Explanations without Opening the Black Box: Automated Decisions and the GDPR. Harvard Journal of Law & Technology 31(2): 841-887.
  20. Rane S (2018) The balance: Accuracy vs. Interpretability.
  21. Sarvakar K, Jani KA, Yagnik SB, Panchal EP, Jain R, et al. (2023) Source-Patent Application Publication India. International classification: B25J 91600: B64C.
  22. Jain R, Vanzara R, Sarvakar K (2024) The Rise of AI and ML in Financial Technology: An In-depth Study of Trends and Challenges. In: Kumar A, et al. (Eds.), Proceedings of the 6th International Conference on Communications and Cyber Physical Engineering ICCCE. Lecture Notes in Electrical Engineering, Singapore 1096: 329-341.
  23. Jain R, Vanzara R (2023) Emerging Trends in AI- Based Stock Market Prediction: A Comprehensive and Systematic Review. Engineering Proceedings 56(1): 254.
  24. Jain R, Mishra S, Prajapati D, Patel C (2023) The Intersection of Artificial Intelligence and Business: A Study of Impact and Implications. Journal of Advances in Artificial Intelligence 1(1): 45-48.
  25. Jain R, Prajapati D, Dangi A (2023) Transforming the Financial Sector: A Review of Recent Advancements in FinTech. International Journal for Research Trends and Innovation 8(2): 250-267.
  26. Patel HB, Mishra S, Jain R, Kansara N (2023) The Future Of Quantum Computing And Its Potential Applications. Journal for Basic Sciences 23(11): 513-519.
  27. Mishra S, Patel HB, Shukla A, Prajapati D, Mevada J, et al. (2023) Call Data Record Analysis using Apriori Algorithm. Indian Journal of Natural Sciences 14(80): 62954-62960.
  28. Jain R, Dhiren P (2023) Climate Change and Animal Health: Impacts, Challenges, and Mitigation Strategies. Open Access Journal of Veterinary Science & Research 8(2): 000249.
  29. Jain R, Dixit S (2023) Revolutionizing the Future: Exploring the Multifaceted Advances of Robotic Technologies. Advances in Robotic Technology 1(1): 1-7.
  30. Bhatla AB, Kikani YB, Joshi DG, Jain R, Patel K (2023) Real Time Cattle Health Monitoring Using IoT, Thing Speak, and a Mobile Application. J Ethol & Animal Sci 5(1): 000131.
  31. Dixit S, Jain R, Patel HB (2024) Impact of 5G Wireless Technologies on Cloud Computing and Internet of Things (IOT). Advances in Robotic Technology 2(1): 1-7.
  32. Mehta D, Patel M, Dangi A, Patwa N, Patel Z, et al. (2024) Exploring the Efficacy of Natural Language Processing and Supervised Learning in the Classification of Fake News Articles. Advances in Robotic Technology 2(1): 7-12.
  33. Jain R, Sarvakar K, Patel C, Mishra S (2024) An Exhaustive Examination of Deep Learning Algorithms: Present Patterns and Prospects for the Future. GRENZE International Journal of Engineering and Technology 10(1): 105-111.
  34. Miller T (2023) Explainable AI: Where Do We Go From Here? Proceedings of the AAAI Conference on Artificial Intelligence 37(1): 333-342.
  35. Ribeiro MT, Singh S, Guestrin C (2024) Beyond Accuracy: Understanding Model Trade-offs in Explainable AI. Nature Machine Intelligence 6(3): 235-245.
More from this journal

Cite this article

BibTeX
APA
RIS
@article{jain2024,
  title   = {Transparency in AI Decision Making: A Survey of Explainable AI
Methods and Applications},
  author  = {Jain R},
  journal = {Advances in Robotic Technology},
  year    = {2024},
  volume  = {2},
  number  = {1},
  doi     = {10.23880/art-16000110}
}
Jain R (2024). Transparency in AI Decision Making: A Survey of Explainable AI
Methods and Applications. Advances in Robotic Technology, 2(1). https://doi.org/10.23880/art-16000110
TY  - JOUR
TI  - Transparency in AI Decision Making: A Survey of Explainable AI
Methods and Applications
AU  - Jain R
JO  - Advances in Robotic Technology
PY  - 2024
VL  - 2
IS  - 1
DO  - 10.23880/art-16000110
ER  -