In order to take the industrial revolution to new horizons (mining and mineral processing), Artificial intelligence and Machine learning can play a critical role. Big data is a fundamental requirement for these techniques, what challenges a researcher has to keep in mind while handling the data coming from industry?
Thaking you all research fellows in advance.
In a more general sense (also applying to your field of mining and mineral processing), there are multiple articles on that topic of general challenges for industrial applications of AI-based systems utilizing Big Data and Machine Learning techniques. The main challenge lies in the fact that there are multiple perspectives from which you can look at it: algorithmic (data science) including preprocessing, database, learning algorithms etc.; acceptance of AI-based systems within an organization including linked topics such as explainability (e.g., explainable AI - XAI), security of such AI-based systems (e.g., adversarial attacks). Depending on the persepctive you are interested in you might want to look up those terms I mentioned above. Furthermore there are some articles that try to summarize some of the challenges (I linked them at the bottom of this answer). Furthermore, you can have a look at our recent papers that deal with decision criteria for AI-based system in an industrial context as well as summary papers on AI security and XAI. I also added a Nature article that reflects on ML application in solid material science that might (or might not) help you out more specifically.
Decision Factors for AI-based systems: Conference Paper How Much AI Do You Require? Decision Factors for Adopting AI...
AI security: Conference Paper Fool me Once, Shame on you, Fool me Twice, Shame on me: A Ta...
Explainable AI: Conference Paper Demystifying the Black Box: A Classification Scheme for Inte...
Papers on challenges:
Conference Paper Big Data for Industry 4.0: A Conceptual Framework
Preprint Industrial Big Data Analytics: Challenges, Methodologies, an...
Article Big Data for Modern Industry: Challenges and Trends [Point of View]
ML in solid material science: Article Recent advances and applications of machine learning in soli...
The noise in the data and its structure, if any, and whether this error structure has time dependance, is a key factor in any data analysis routine.
The big problem is reliable fast algorithms. Although, recently deep learning algorithms have a boom, but they are never reliable and seems to be guesses/forcibly compelled to optimality.
- Badges
- Science topic
- Similar topics
- Computer Science
- Data Mining