Posts from March 2019 (15)

We’re three weeks into the month-long Ihaka Lecture Series, and it has been well received – thank you to those who have turned up in person and online.

Our final speaker, Robert Tibshirani, right, is up on Weds April 3 at the University of Auckland (details here). Robert is Professor of Statistics and Biomedical Data Science at Stanford University.

He is best known for proposing the ‘lasso’, a sparse regression estimator, and describing its relationship to the idea of boosting in supervised classification. He will talk about modern sparse supervised learning approaches that extend the lasso.

In the meantime, you might like to check out the films of the last three speakers. First up on March 13 was by Bernhard Pfahringer, left, who is Professor of Computer Science at the University of Waikato.

He is a member of the Weka project, New Zealand’s other famous open-source data science contribution, and here talks about the design and development of Weka and more recent projects.

Next was supposed to be JJ Allaire, the founder and CEO of RStudio, and the author of R interfaces to Tensorflow and Keras. However, ill-health prevented him coming, and our very own Professor Thomas Lumley stepped in.

Thomas talked entertainingly about deep learning, in particular how deep convolutional nets are structured and how they can be remarkably effective, but can also fail, as he puts it, “in remarkably alien ways”.

Following was Dr Kristian Lum, Lead Statistician at the Human Rights Data Analysis Group. Her research has concretely demonstrated the potential for machine learning-based predictive policing models to reinforce and, in some cases, amplify historical racial biases in law enforcement.

She talked about algorithmic fairness, and about ways in which policy, rather than data science, influences the development of these models and their choice over non-algorithmic approaches.

Women all know about the toilet queue in the intermission at concerts – same-sized bathrooms for men and women does not equal efficiency. Women who have ever stood and waited in a long line for the loo while the men come and go with speed – and I think I can say that this is about, roughly, give or take, 100% of us – roll our eyes and laugh about this as we wait. But the anecdote reveals an uncomfortable truth, says Caroline Criado Perez in her book Invisible Women: Exposing Data Bias in a World Designed for Men. Design and services that takes the average male or the needs of the average male as the norm – as is the case with car-crash test dummies and stab-proof vests, among other things – are potentially deadly. The Guardian has excerpted a section of her book and it’s a sobering read. Recommended.

And while we are on the subject of a world designed for data about men, NASA has cancelled the first all-women spacewalk due to a spacesuit size issue.

Team Ratings for Round 7

The basic method is described on my Department home page.
Here are the team ratings prior to this week’s games, along with the ratings at the start of the season.

Performance So Far

So far there have been 40 matches played, 26 of which were correctly predicted, a success rate of 65%.
Here are the predictions for last week’s games.

Predictions for Round 7

Here are the predictions for Round 7. The prediction is my estimated expected points difference with a positive margin being a win to the home team, and a negative margin a win to the away team.

Team Ratings for Round 18

The basic method is described on my Department home page.
Here are the team ratings prior to this week’s games, along with the ratings at the start of the season.

Performance So Far

So far there have been 102 matches played, 70 of which were correctly predicted, a success rate of 68.6%.
Here are the predictions for last week’s games.

Predictions for Round 18

Here are the predictions for Round 18. The prediction is my estimated expected points difference with a positive margin being a win to the home team, and a negative margin a win to the away team.

Team Ratings for Round 19

The basic method is described on my Department home page.
Here are the team ratings prior to this week’s games, along with the ratings at the start of the season.

Performance So Far

So far there have been 126 matches played, 100 of which were correctly predicted, a success rate of 79.4%.
Here are the predictions for last week’s games.

Predictions for Round 19

Here are the predictions for Round 19. The prediction is my estimated expected points difference with a positive margin being a win to the home team, and a negative margin a win to the away team.

Team Ratings for Round 3

The basic method is described on my Department home page.
Here are the team ratings prior to this week’s games, along with the ratings at the start of the season.

Performance So Far

So far there have been 16 matches played, 11 of which were correctly predicted, a success rate of 68.8%.
Here are the predictions for last week’s games.

Predictions for Round 3

Here are the predictions for Round 3. The prediction is my estimated expected points difference with a positive margin being a win to the home team, and a negative margin a win to the away team.

Clinical science stories are the opposite of most current events coverage: the good news is dramatic and has an army of publicists; the bad is slow and boring.

In September 2016 there was a flurry of news articles about a new candidate treatment for Alzheimer’s

In the Herald (from the Telegraph)

• Hope for drug to beat Alzheimer’s. Scientists show they can halt mental decline by cleaning brain of sticky plaque.

(from the Washington Post)

• A glimmer of hope for an Alzheimer’s drug. An initial trial of an antibody therapy that targets Alzheimer’s disease has shown promising results and could signal a long-awaited breakthrough in treating the devastating brain disorder that affects millions.

Stuff (from the Telegraph)

• Alzheimer’s drug could be ‘game changer’. The first drug to combat Alzheimer’s disease is on the horizon after scientists proved they can halt mental decline by clearing the sticky plaques from the brain that cause dementia.

TVNZ

• Alzheimers treatment drug trial gives researchers hope. A drug which reduces the amount of protein plaque – called amyloid – in the brain is giving hope to researchers and Alzheimers sufferers alike.

Radio NZ

• Science: Alzheimer’s breakthrough. A new drug that seems to stop the progress of Alzheimer’s Disease has been tested on patients in America.

Now, in 2019, two large clinical trials were just stopped because the treatment didn’t work.  This sort of thing happens a lot; drug development is hard, especially when we don’t actually understand the disease very well.

The coverage of Alzheimer’s treatment candidates has two big problems, though. First, promising initial results are almost always oversold. Second, the failures usually aren’t covered, except perhaps in the business pages.

Team Ratings for Round 6

The basic method is described on my Department home page.
Here are the team ratings prior to this week’s games, along with the ratings at the start of the season.

Performance So Far

So far there have been 33 matches played, 23 of which were correctly predicted, a success rate of 69.7%.
Here are the predictions for last week’s games.

Predictions for Round 6

Here are the predictions for Round 6. The prediction is my estimated expected points difference with a positive margin being a win to the home team, and a negative margin a win to the away team.

Team Ratings for Round 2

The basic method is described on my Department home page.
Here are the team ratings prior to this week’s games, along with the ratings at the start of the season.

Performance So Far

So far there have been 8 matches played, 5 of which were correctly predicted, a success rate of 62.5%.
Here are the predictions for last week’s games.

Predictions for Round 2

Here are the predictions for Round 2. The prediction is my estimated expected points difference with a positive margin being a win to the home team, and a negative margin a win to the away team.

Team Ratings for Round 5

The basic method is described on my Department home page.
Here are the team ratings prior to this week’s games, along with the ratings at the start of the season.

Performance So Far

So far there have been 28 matches played, 19 of which were correctly predicted, a success rate of 67.9%.
Here are the predictions for last week’s games.

Predictions for Round 5

Here are the predictions for Round 5. The prediction is my estimated expected points difference with a positive margin being a win to the home team, and a negative margin a win to the away team.