- Uplift Models with Multiple Treatments: A well defined/motivated problem that extends uplift models to a larger domain. Good presentation.

- MultiC: A problem of expanding a model for one crowd sourced problem (e.g. predict wild cats) with a similar one (predict wild dogs); neat

- HiDDen: The basic method is well-motivated, but I'm curious if there is a more mathematical way to validate the model utility.

- Next Term Grade Prediction: Would be interesting to consider course load attributes or extend the system to select course combinations.

- BreachRadar: Interesting, but I think it is unfortunate to not use some sort of company derived baseline of cost savings.

effective, considering variability in how connections are formed based on individual experience. Maybe certain processes are more stable.

Latent Semantic Factors: A very well motivated problem of inferring expert reviewers and their development to better model helpful reviews.

considering different word meanings etc. Integrates in user's tendency toward pessimism/optimism. Neat!

User Guided Cross-Domain Sentiment: A new method of inferring the goodness of reviews in one domain (e.g. books) to another (e.g. music)...

... Wish there were a few more examples of applying the concept though!

- Seriation Presentation: Interesting application of ordering instances the most consistently. Benefit for larger data sets is impressive...

Land Cover Prediction: Methods to predict changing landcover from satellites. Surprising that it doesn't require specific domain knowledge.

Generalized Inverse Classification: Cool work on ``adjusting'' attributes to improve outcomes for general classification models.

Active Cost-sensitive -> expected error: An important refinement of active learning. Not my area so I can't add much more than that!

Power of Certainty: The authors introduce a more representative, smoother way to characterize social media influence. Cool!

to pick. Been a while since I've read that sort of material so I can't comment too deeply about it, unfortunately!