Emerging Technologies (AI, Cloud, IoT, Big Data)
Core concepts
Artificial intelligence spans machine learning (supervised, unsupervised, reinforcement), neural networks and deep learning, natural language processing, and expert systems. Know the difference between training and inference, overfitting versus underfitting, and evaluation metrics (accuracy, precision, recall, F1).
Cloud computing covers service models (IaaS, PaaS, SaaS) and deployment models (public, private, hybrid, community), plus characteristics such as on-demand self-service, elasticity, and measured service. Know virtualization and containers.
IoT describes sensor-actuator networks connected to the internet, with edge and fog computing reducing latency; consider security and privacy risks. Big Data is characterized by volume, velocity, variety, veracity, and value, processed by frameworks such as Hadoop (HDFS, MapReduce) and Spark. Analytics tiers are descriptive, diagnostic, predictive, and prescriptive.
Traps to watch
- Confusing IaaS (infrastructure) with PaaS (platform) and SaaS (software).
- Assuming more data always improves a model; quality matters.
- Mixing precision (of predicted positives, how many are correct) with recall (of actual positives, how many were found).
- Treating IoT devices as inherently secure; they are often constrained.
Worked example
A classifier flags 20 emails as spam; 15 are truly spam. Of 30 actual spam emails, it found 15. Precision = 15/20 = 0.75; recall = 15/30 = 0.50. Rule of thumb: precision penalizes false positives; recall penalizes false negatives.