The Hark Blog

Splitting data

Splitting data into ‘train’, ‘validation’ and ‘test’ sets

When developing and deploying machine learning models, it’s important that we split the dataset into ‘train’, ‘validation’, and ‘test’ datasets. This protects against an overfitted model, and helps ensure results are generalised. In this blog post we will look in to how to split the data, and why.

project management

8 Project management methodologies for delivering IoT deployments

Internet Of Things (IoT), is no longer a tech buzzword but an advanced everyday reality describing a multitude of technologies that connect assets and objects to the internet. Before you start deploying IoT solutions or deployments, it’s vital to understand what project management methodology to use, as all IoT projects are different and will require slightly different approaches.

puzzle piece missing - missing data

How to deal with missing data

In Data Science, it’s pretty rare to come across a perfect dataset; data will frequently be of poor quality. It could have been recorded incorrectly, or it can be missing completely, which can be caused by poor data collection or storage methodologies.

Phoebe leading discovery sessionj

Finding the solution means defining the problem

The move towards Net Zero, for companies big or small, can be really daunting. Where do you start? How much will it cost? What’s the ROI? Who can help me? These are all reasonable questions that create obstacles to a net zero deployment.

energy ball of electricity

Hark Energy Hierarchies; the most useful features for data validation

As of 2022, there hasn’t been a more vital time to have a deeper understanding of your energy consumption; whether that’s across a varied estate of sites, in a factory, depot, or store. The fact is that energy costs are exponentially going up, without any indication of whether they’ll come back down. The time for action is now.

Subscribe to Our Blog

Stay up to date with the latest industry news, platform developments and more.