ST⛔P assuming data science helps make better or smarter decisions with ai ml LLMs data and related technologies and algorithms development.

We should transform the data science approach into something meaningful for business to help organizations make better business decisions.

Why? Let's have a look at the high-level process for most data science teams at the moment.
1️⃣ Problem analysis and understanding
2️⃣ Data acquisition and exploration
3️⃣ Algorithm (model) development: algo selection, optimization, training, validation etc.
4️⃣ Deployment and monitoring

This is a flawed process. I talked about confirmation bias multiple times, and that is going to happen in step 2. Even if the team's lottery wins and they build a relatively good ai/ml model, more often than not (a) they don't leave the development phase to production (complexity of operationalizing, integration, changing org behavior, and so on). Or (b) when they do leave the development phase, businesses do not adopt it because that's not what the business had needed in the first place!

As we see from the above, there are 3 outcomes for the data-centric approach:
🚨Bias models (because of human confirmation bias)
🚨Not leaving the development phase (complexity of integration, operationalizing, accuracy, etc.)
🚨Solving the wrong problem (type 3 error) datascientist

As we see, the definite outcome of the current process and status quo in the data-driven approach for DataScience is an absolute failure.

🤔So what is the solution?
1️⃣ Start with the business objectives, but then do NOT jump to “data acquisition and exploration”
2️⃣ Frame the “business decisions” that have an impact on those objectives by decomposing decisions into smaller, more understandable decision units.
3️⃣ Depict the dependencies and relation between them and identify how they impact the business objectives in the context of the problem scope.
4️⃣ Determine the requirements for each decision unit (i.e. appropriate techniques and technology, algorithms, data, domain knowledge, rules etc.)

Now you have the blueprint of moving the needle in the right direction to “Improve the Speed and Quality of Key Business Decisions” related to the business objectives that matter to business.

🎯This is how the Decision-Centric Approach works and will guarantee to delivery of business value to organizations. It should not be a guesswork. It's a science, not a data science, the decision science!

Learn more at https://lnkd.in/gcWUy9eU
–
Follow me if you are looking for unfiltered views into decisionintelligence ai. Hit 🔔on my profile and get a notification for all my new daily posts.

Posted here.

Published November 28th, 2024 at 07:30 am