The Ledger

Data Quality Directly Affects the Quality of Your Predictive Model

“Poor data quality is enemy number one to the widespread, profitable use of machine learning. “

Data is absolutely key to the success of any machine learning process. If there is bad data in the process, the bad effects will be seen twice-  first in the historical data used to train the predictive model and second in the new data used by that model to make future decisions. The recipe for a strong predictive model is accurate, quality data and having the right data. The right data for the model is lots of unbiased data, over the entire range of inputs for which one aims to develop the predictive model. Most data quality work focuses on one criterion or the other, but for machine learning, you must work on both simultaneously.

Read More at The Harvard Business Review >

Management Accountants: Succeeding in The Digital Age

“Having a view of a business structured according to financial reporting rules and standards is important. Yet having accounting systems that deliver only such information is insufficient for providing the guidance enterprises need to succeed. In order to succeed in today’s increasingly competitive environment, organizations will need to understand the need for cost modeling that adequately supports its managers’ decision-making requirements.”

Read More at Strategic Finance Magazine >

The Power of Data Should Never be Underestimated

Data is the foundation for any successful business model. We can now embed and extract data and intelligence out of every business transaction and physical object, from machines and cars to wearables, and we can harness the power of this connected data to create new value in the form of unprecedented revenue opportunities, efficiency improvements, and better customer experiences. Often, companies already have the key ingredients for using data to their advantage, though they may not be aware of the “how.”

Read More at The Digitalist by SAP >

Transparency Through OEMs is Key for Manufacturers

As manufacturers respond to new competitive and consumer demands, they are relying on OEMs (original equipment manufacturers) to take on more responsibility, as well. OEMs are manufacturers who resell another company’s product under their own name and branding. To provide visibility, OEMs need more information and visibility into production schedules, constraints, changeover frequency and the overarching expectations of the production lines to gain insight into the factory’s culture and infrastructure. However, what’s missing in the factory setting is an enterprise view that looks at the entire production facility and applies predictive analytics to what should happen if there is an issue in a specific area of the plant.

Read More at Automation World >