Thinking big at Yahoo

I’m speak­ing in the “Yahoo Labs Big Thinkers” series on Fri­day 26 June. I hope I can live up to the title!

My talk is on “Explor­ing the bound­aries of pre­dictabil­ity: what can we fore­cast, and when should we give up?”  Essen­tially I will start with some of the ideas in this post, and then dis­cuss the fea­tures of hard-​​to-​​forecast time series.

So if you’re in the San Fran­cisco Bay area, please come along. Oth­er­wise, it will be streamed live on the Yahoo Labs web­site. Con­tinue reading →

Travelling Thilaksha

One of my PhD stu­dents, Thi­lak­sha Tha­ranganie, has been very suc­cess­ful in get­ting travel fund­ing to attend con­fer­ences. She was the sub­ject of a write-​​up in today’s Monash News.

We encour­age stu­dents to attend con­fer­ences, and pro­vide fund­ing for them to attend one inter­na­tional con­fer­ence and one local con­fer­ence dur­ing their PhD can­di­da­ture. Thi­lak­sha was pre­vi­ously funded to attend last year’s COMPSTAT in Geneva, Switzer­land and IMS con­fer­ence in Syd­ney. Hav­ing exhausted local fund­ing, she has now con­vinced sev­eral other orga­ni­za­tions to sup­port her con­fer­ence habit.

Now she just has to fin­ish that thesis…

Paperpile makes me more productive

One of the first things I tell my new research stu­dents is to use a ref­er­ence man­age­ment sys­tem to help them keep track of the papers they read, and to assist in cre­at­ing bib files for their bib­li­og­ra­phy. Most of them use Mende­ley, one or two use Zotero. Both do a good job and both are free.

I use nei­ther. I did use Mende­ley for sev­eral years (and blogged about it a few years ago), but it became slower and slower to sync as my ref­er­ence col­lec­tion grew. Even­tu­ally it sim­ply couldn’t han­dle the load. I have over 11,000 papers in my col­lec­tion of papers, and I was spend­ing sev­eral min­utes every day wait­ing for Mende­ley just to update the database.

Then I came across Paper­pile, which is not so well known as some of its com­peti­tors, but it is truly awe­some. I’ve now been using it for over a year, and I have grown to depend on it every day to keep track of all the papers I read, and to cre­ate my bib files. Con­tinue reading →

What to cite?

This ques­tion comes from a com­ment on another post:

I’ve seen authors cit­ing as many ref­er­ences as pos­si­ble to try to please poten­tial ref­er­ees. Many of those ref­er­ences are low qual­ity papers though. Any gen­eral guid­ance about a typ­i­cal length for the ref­er­ence section?

It depends on the sub­ject and style of the paper. I’ve writ­ten a paper with over 900 cita­tions, but that was a review of time series fore­cast­ing over a 25 year period, and so it had to include a lot of references.

I’ve also writ­ten a paper with just four cita­tions. As it was a com­men­tary, it did not need a lot of con­tex­tual information.

Rather than pro­vide guid­ance on the length of the ref­er­ence sec­tion, I think it is bet­ter to fol­low some gen­eral prin­ci­ples of cita­tion in research. Con­tinue reading →

Dark themes for writing

I spend much of my day sit­ting in front of a screen, cod­ing or writ­ing. To limit the strain on my eyes, I use a dark theme as much as pos­si­ble. That is, I write with light col­ored text on a dark back­ground. I don’t know why this is not the default in more soft­ware as it makes a big dif­fer­ence after a few hours of writing.

Most of the time, I am writ­ing using either Sub­lime Text, RStu­dio or TeX­studio. Each of them can be set to use a dark theme with syn­tax col­or­ing to high­light struc­tural fea­tures in the text.
Con­tinue reading →