Augmenting Organizational Decision-Making with Deep Learning Algorithms: Principles, Promises, and Challenges

Yash Raj Shrestha,Vaibhav Krishna,Georg von Krogh

DOI: https://doi.org/10.48550/arXiv.2011.02834

2020-11-02

Abstract:The current expansion of theory and research on artificial intelligence in management and organization studies has revitalized the theory and research on decision-making in organizations. In particular, recent advances in deep learning (DL) algorithms promise benefits for decision-making within organizations, such as assisting employees with information processing, thereby augment their analytical capabilities and perhaps help their transition to more creative work.

Machine Learning

What problem does this paper attempt to address?

The core problem that this paper attempts to solve is: how to use deep learning (DL) algorithms to enhance the decision - making process of organizations, thereby improving information processing efficiency and decision - making quality. Specifically, the authors explored the application potential, advantages, and challenges of deep learning in organizational decision - making.

### Problem Background

1. **Traditional Challenges in Organizational Decision - making**:
- Organizational decision - making has long relied on effective information processing capabilities to ensure the quality and efficiency of decisions.
- With the development of information technology, especially the progress of artificial intelligence (AI) and machine learning (ML), deep learning algorithms provide new tools for solving complex information processing tasks.

2. **Advantages of Deep Learning**:
- Deep learning algorithms can extract patterns from a large amount of unstructured data and make accurate predictions, which is of great significance for marketing, finance, operations, human resources, and strategic decision - making within the organization.
- The decision - support system enhanced by deep learning (DLADM) can significantly improve employees' information processing and analytical capabilities, helping them transform to more creative tasks.

### Main Contributions of the Paper

1. **Conceptual Framework**:
- Proposed the concept of "deep learning - enhanced decision - making" (DLADM), that is, enhancing the decision - making process of organizations through the output of deep learning algorithms (such as predictions or robust patterns extracted from unstructured data).

2. **Technology Introduction**:
- Provided an easy - to - understand tutorial on deep learning algorithms to help readers understand how these algorithms work and their performance in practical applications.

3. **Case Studies**:
- Demonstrated the application effects of DLADM through two specific case studies (image recognition and sentiment analysis). These two cases used datasets from the European e - commerce company Zalando and the movie review aggregation website Rotten Tomatoes respectively.

4. **Challenges and Suggestions**:
- Discussed the economic and organizational challenges faced by DLADM in practical applications, including the costs of data collection and labeling, the requirements for computing infrastructure, and risks in terms of fairness, responsibility, reliability, and trust.
- Proposed some suggestions to help managers deal with these challenges and successfully introduce deep learning technology into their organizations.

### Conclusion

This paper aims to provide a comprehensive understanding framework for scholars and practitioners by in - depth exploration of the application of deep learning in organizational decision - making, promoting broader adoption and optimization of deep learning technology to enhance the overall decision - making ability and performance of organizations.