O'Reilly on Data

DECEMBER 12, 2019

Not least is the broadening realization that ML models can fail. And that’s why model debugging, the art and science of understanding and fixing problems in ML models, is so critical to the future of ML. Because all ML models make mistakes, everyone who cares about ML should also care about model debugging. [1]

Machine Learning 281

Machine Learning Modeling Testing Risk Management 281

Smart Data Collective

SEPTEMBER 17, 2020

A 2013 survey conducted by the IBM’s Institute of Business Value and the University of Oxford showed that 71% of the financial service firms had already adopted analytics and big data. Big Data can efficiently enhance the ways firms utilize predictive models in the risk management discipline. The Underlying Concept.

Big Data 145

Big Data Finance Risk Management Risk 145

Domino Data Lab

AUGUST 22, 2019

The excerpt covers how to create word vectors and utilize them as an input into a deep learning model. While the field of computational linguistics, or Natural Language Processing (NLP), has been around for decades, the increased interest in and use of deep learning models has also propelled applications of NLP forward within industry.

Deep Learning 111

Deep Learning Modeling Metrics Testing 111

Domino Data Lab

APRIL 21, 2021

We’ll use a gradient boosting technique via XGBoost to create a model and I’ll walk you through steps you can take to avoid overfitting and build a model that is fit for purpose and ready for production. This is generally problematic, as the model trained on such data will have difficulties recognising the minority class.

Statistics 69

Statistics Machine Learning Modeling Metrics 69

DMBS2

MAY 20, 2018

Multiple kinds of data model are viable … … but it’s usually helpful to be able to do some kind of JOIN. The ongoing rise of “edge computing” and the “Internet of Things” fit into the general trend that in 2013 I summarized as appliances, clusters and clouds.

Internet of Things 50

Internet of Things Predictive Analytics Metrics Enterprise 50

Domino Data Lab

JULY 9, 2019

Assessing whether a business stakeholder is trying to solve for a problem that is descriptive, predictive, or prescriptive and then re-framing the problem as supervised learning, unsupervised learning, or reinforcement learning, respectively. One of the ways I frame that is, “Are you looking to build a predictive model?

Data Science 112

Data Science Machine Learning Advertising Modeling 112

Jet Global

MAY 1, 2023

Companies like Tableau (which raised over $250 million when it had its IPO in 2013) demonstrated an unmet need in the market. These licensing terms are critical: Perpetual license vs subscription: Subscription is a pay-as-you-go model that provides flexibility as you evaluate a vendor. Their dashboards were visually stunning.

Analytics 52

Analytics Cost-Benefit Visualization Dashboards 52