Level Up Your Machine Learning Skills: 5 Must-Watch Lectures

Machine learning (ML) is revolutionizing various fields, but it's true power lies in its ability to scale. This curated list of lectures dives deep into specific, yet valuable, areas within ML, equipping you with the knowledge to tackle complex challenges.

Whether you're a seasoned data scientist or just starting your ML journey, these lectures offer something for everyone. From practical tools for anomaly detection to the ethical considerations of predicting police misconduct, you'll gain insights and techniques to elevate your skillset.

1. MLOps so AI can scale. Most of the failures in ML development come not from developing poor models but from poor productization practices (McKinsey).

2. How to Detect the Trend in the Time Series Data and Detrend in Python (Regenerative).

3. Predicting Police Misconduct: Whether police misconduct can be prevented depends partly on whether it can be predicted. We show police misconduct is partially predictable and that estimated misconduct risk is not simply an artifact of measurement error or a proxy for officer activity. We also show many officers at risk of on-duty misconduct have elevated off-duty risk too, suggesting a potential link between accountability and officer wellness. We show that targeting preventive interventions even with a simple prediction model – number of past complaints, which is not as predictive as machine learning but lower-cost to deploy – has marginal value of public funds of infinity (NBER).

4. Synthetic controls with machine learning: application on the effect of labour deregulation on worker productivity in Brazil. Synthetic control methods are a data-driven way to calculate counterfactuals from control individuals for the estimation of treatment effects in many settings of empirical importance. In canonical implementations, this weighting is linear and the key methodological steps of donor pool selection and covariate comparison between the treated entity and its synthetic control depend on some degree of subjective judgment. Thus current methods may not perform best in settings with large datasets or when the best synthetic control is obtained by a nonlinear combination of donor pool individuals. This paper proposes "machine controls", synthetic controls based on automated donor pool selection through clustering algorithms, supervised learning for flexible non-linear weighting of control entities and manifold learning to confirm numerically whether the synthetic control indeed resembles the target unit. The machine controls method is demonstrated with the effect of the 2017 labour deregulation on worker productivity in Brazil. Contrary to policymaker expectations at the time of enactment of the reform, there is no discernible effect on worker productivity. This result points to the deep challenges in increasing the level of productivity, and with it, economic welfare (BIS).

5. Finding a needle in a haystack: a machine learning framework for anomaly detection in payment systems. We propose a flexible machine learning (ML) framework for real-time transaction monitoring in high-value payment systems (HVPS), which are a central piece of a country's financial infrastructure. This framework can be used by system operators and overseers to detect anomalous transactions, which - if caused by a cyber attack or an operational outage and left undetected - could have serious implications for the HVPS, its participants and the financial system more broadly. Given the substantial volume of payments settled each day and the scarcity of actual anomalous transactions in HVPS, detecting anomalies resembles an attempt to find a needle in a haystack. Therefore, our framework uses a layered approach. In the first layer, a supervised ML algorithm is used to identify and separate 'typical' payments from 'unusual' payments. In the second layer, only the 'unusual' payments are run through an unsupervised ML algorithm for anomaly detection. We test this framework using artificially manipulated transactions and payments data from the Canadian HVPS. The ML algorithm employed in the first layer achieves a detection rate of 93%, marking a significant improvement over commonly-used econometric models. Moreover, the ML algorithm used in the second layer marks the artificially manipulated transactions as nearly twice as suspicious as the original transactions, proving its effectiveness (BIS).

So, dive in, explore these lectures, and embark on a path of continuous learning in the ever-evolving world of machine learning! Remember, the key to success in AI lies not just in building models, but in building a robust and scalable ML ecosystem. Happy learning!

Feel free to follow me on X