6

Peter's Personal Review for Oct-Dec 2014

Happy 2015 everyone!

Last year, I did three month personal reviews (Dec '13 - Feb '14, March-May '14, May-June '14, and July-Sep '14) to track my progress against my EA goals. This is definitely something that I want to keep doing, and I think the three month cycle works well (after having tried reviews on one-month and six-month cycles). In doing this review, I hope to be not all that different from the three month review of any other EA org, except I'm an EA org of precisely one person.

Hopefully, posting such a review here is useful as a case-study of a particular effective altruist approach, and you might find something useful in it, or have some useful commentary to offer me.

 

How Did I Spend My Time?

These three months, I used Toggl for to-the-minute timetracking, though there are some inaccuracies.


Total

Per Week[1]

% Week

Sleep

767hrs

59.0hrs

35.1%

Social

379

29.2

17.4%

Day Job

331

25.5

15.2%

Break

181

14.0

8.3%

EA Direct

147

11.3

6.7%

Prep[2]

119

9.2

5.4%

Planning

67

5.0

2.9%

Other

58

4.5

2.7%

Entrepreneurship

45

3.5

2.1%

Email

38

3.0

1.8%

Cleaning

29

2.2

1.3%

Programming

23

1.8

1.1%

Food[3]

19

1.5

0.8%

Chinese

19

1.5

0.8%

Exercise

18

1.4

0.8%

Errands

13

1.0

0.6%

Writing[4]

6

0.5

0.3%

Moving

0

0.0

0.0%

[1]: On average – obviously not all of these activities took place every week.
[2]: Includes getting ready in the morning and all commuting to and from places (not just work).
[3]: Includes making, but not eating food.
[4]: Does not include writing for the EA Blog, which is considered “EA Direct Work”.

 

Day Job

I work as a data scientist at a fast-growing start-up in Chicago.

 

Day Job: What Went Right

At the end of my last review, I had just finished my internship as a software engineer and received a permanent position. I was offered to either be a software engineer or a data scientist, and I chose to become a data scientist, because I thought the career was more intrinsically interesting and data science seems to me to be more likely to have some staying power if it turns out we’re in another tech bubble (which seems quite plausible to me).

I was excited by this transition because while I felt like I had worked a lot with statistics in college, I didn’t know a lot about how it worked in the real world, where the statistical models would be under a lot of demand. I learned a lot about “data engineering”, which is the work of moving data around and building data pipelines that can work under heavy load. I spent much more of my job doing that, and actually haven’t learned much data science yet.

While I feel like I learned a lot about R in my last review (July-Sep ‘14), I definitely didn’t know as much about it as I did now. I feel pretty confident now in both R and Ruby, and can now look back on some of the code I wrote July-Sep ‘14 and laugh at how much of a beginner I once was. I think I got my first tastes of “advanced” R scripting in Oct-Dec ‘14, which culminated in writing the plugin Batchman for batching processes in R. I’m excited to learn more!

I also got my first taste of “big data”, where the amount of data I was working with was too big for R to handle all at once. This is new for me, and also pretty new for our start-up. There are definitely more mainstream tools to be used here (e.g., Hadoop), but I took the approach of trying to build some tools for R to work with big data instead.

Lastly, I took the chance to work on some different skills and picked up more responsibilities by becoming a part-part-time project manager for our team. Most of the teams here have their own full-time project managers, but data science was too domain heavy to make it easy to recruit one. I volunteered to fill that gap. So I’ve been applying organizational skills to make sure our team is on-track.


Overall, I really like my co-workers, I feel like I’m learning a lot, I feel like my work is high challenge while still being possible, and I really like our work’s flexible schedule to allow me to work on EA projects when I want to.

I’ve also been excited by the salary afforded to me by this job, which has allowed me to make some donations, noted in my donation log.

 

Day Job: What Went Wrong

I do feel bad about the 25.5hrs a week figure.

I can defend it, of course. First, I’m using an actual timer to determine how much time I actually work, and I’m not counting going out for lunch or even going to the bathroom. If you took a typical 40hr/wk person and subtracted out their Reddit time and bathroom time and lunch hours, I’m sure they would be much lower than 40 hours.

Second, there was an awful lot of holidays in November and December, and I took two weeks off.

Third, I feel like data science, and anything with software really, is truly a “work smarter, not harder” kind of industry where the benefits of your results frequently are not a result of how many hours you spent working on them.

And fourth, maybe the utilitarian in me should be glad that I’m getting the same salary, but am able to put more hours into EA projects?

But all that being said, I think it would be beneficial to me personally if I spent more work on my job, because I think I would learn more in response. Maybe I can start working on learning more parts of the data science industry, such as modeling or devops. ...Or I could get better at big data, such as actually learning how to use Hadoop. I imagine these skills would enhance my future career prospects pretty well.

 

Day Job: Goals to Accomplish by Next Review

  • Get back up to spending 30hrs/wk on my day job.

  • Learn some data modeling.

  • Stretch goal -- Learn Hadoop.

 

EA Direct Work

I continue to maintain a large medley of EA projects. I think this is smart, because the only way I can reliably know whether they’ll work is to actually do them -- I’m not really able to tell in advance.

To make this a bit more clear, here’s a breakdown:


Total

Per Week

% EA

EA Survey

56.8hrs

4.4hrs

38.6%

Charity Science

22.6

1.7

15.4%

Birthday for Charity

18.5

1.4

12.6%

Pro-veg Research

13.8

1.1

9.4%

Blogging

11.4

0.9

7.7%

.impact

10.6

0.8

7.2%

Other

13.3

1

9.1%


EA Direct: What Went Right

Charity Science -- I’m pleased to be an early member of Charity Science. I’ve been ramping up my involvement with them lately, so the small hours per week isn’t really indicative of more recent trends. I’ll wait for the official Charity Science three month review to explain our success in more detail, but we’ve definitely moved over $100K in 2014 to GiveWell’s top charities.

My personal Birthday for Charity fundraiser -- I raised over $5K for AMF! I explain how that happened elsewhere. Giles took those lessons, applied them diligently, and raised over $2K himself. Not bad!

Pro-veg research -- Once again, I was given the opportunity to play with far more money than I earned this year, when Jason and I teamed up with Mercy for Animals to run our study into vegetarian advocacy. The biggest success here is learning delegation, by delegating a lot to Jason and to MFA -- things they would do better than I would! I also managed to delegate some other veg research, such as a look into Humane Spot and a study on disguising the intent of vegetarian questionnaires (forthcoming).

I worked with Tom to add a feature to this EA forum, the "More Comments" link underneith comments in the sidebar.  If you saw how complex the EA Forum codebase is, you'd be impressed too.

 

EA Direct: What Went Wrong

Survey Data Analysis -- This has been a lesson in the planning fallacy. During my last review, I was convinced that the EA Survey was right around the corner. I was wrong -- I spent nearly 60hrs on it to finish, way more than I expected it would take, but the diligence was important. But then before I could get the final touches done, I hit Thanksgiving, my birthday fundraiser, Christmas, and then New Years, and had no time to get it out the door. It’s still sitting there. I’m disappointed to not have this done by this review, but I’ll definitely have it done by the next one.

Chicago Skeptics Meetup -- The Chicago Skeptics have two meetups a month. I went to one in September and promised I’d be back to the next one. I never returned… to that meetup, or any of the ones over the next four months. I regret that -- I think the networking potential is high, the returns from further testing Charity Science’s modeling of targeting skeptics is high, and it was just plain good for me socially.

 

EA Direct: Goals to Accomplish by Next Review

  • Publish the EA Survey.

  • Attend at least one more Chicago Skeptics meetup.

  • Continue to work with Charity Science to test more fundraising ideas, ramping up involvement.

  • Continue to advise MFA on the veg study, but decrease involvement.

  • Decrease involvement in other EA activities and time spent on EA direct work overall.

 

Entrepreneurship

I’m still trying to work on some projects to try to bring in money, whether they take the form of “passive income” projects or potential start-up ideas. I have a long-term goal of eventually replacing my day job income and either (a) making an income cushion so I can do EA projects full-time or (b) creating a rather successful (>$5M in valuation) company.

Where am I at this goal? September saw the launch of 64or32.com, which has about $6 in recurring monthly revenue (from ads), and which I’d value at about $100. Still have a few more zeroes to tack onto that number before I feel successful. But at least it’s paid for itself, and I’m no longer in the red!

My business partner and I did a lot of work behind the scenes to bring some other ideas to the MVP stage, and I think we have four potential projects ready to see if they’ll get us any money.

 

Entrepreneurship: What Went Right

The biggest thing I’m excited about so far is our idea generation -- it’s greatly surpassed our ability to actually execute, and it means that if our first couple of ideas inevitably fail, we’ll have a lot to try next!

 

Entrepreneurship: What Went Wrong

I’m still disappointed we didn’t actually launch something.


Entrepreneurship: Goals to Accomplish by Next Review

  • Launch something.

  • Stretch goal -- launch all four of the projects we think are ready to launch soon.

  • Stretch goal -- double monthly revenue (sounds impressive until you remember that’s bringing us from $6/mo to $12/mo).

 

Other Things

Chinese -- I’m trying to learn Chinese now. It’s pretty new -- a goal I started in December. I’m mainly doing this because it’s fun, though I think there are benefits to learning it. It seems like a generally good skill to have.

Programming -- I spent some time outside of work honing my programming skills. I wroteGit it on, a neat plugin for quickly navigating GitHub from the command line. I wrote Git-aliases, which is a compilation of aliases for interacting with GitHub more quickly. I also wrote some more R plugins, such as SummarizeR, which generates summary reports of data, and Surveytools, a modern, lightweight workflow for analyzing survey data. I’d love to work on these more and get them up on CRAN and Hacker News (goals for the next review).

Exercise -- I’m disappointed that I’m not taking exercise seriously enough, given the large effect it has on happiness, productivity, and well-being (all three things I like). I didn’t exercise much Jul-Sep ‘14, and I wanted to turn that around for Oct-Dec ‘14. I succeeded in exercising more (0.4hrs/wk -> 1.4hrs/wk), even buying a pull-up bar, but I’m still not where I’d like to be. My exercising is still inconsistent, and my weight at 31 Dec 2014 was heavier than it was at 1 Jan 2014. I think as much as I don’t like the idea, I need to pony up the money and go to an actual gym and lift actual weights, and double my time spent exercising per week.  (Update 3 Jan - going to try to stick to this exercise plan.)

Social -- I like to socialize with people. This number is high because I had about 20 days (>20% of the review period) where I was on vacation, because our civilization decided to smush the year’s three most important holidays together into a block of 36 days. I expect this number to go down for the next review period.

Sleep and Break -- These numbers are about as good as they’re going to get, and attempting to decrease them further will only be counterproductive. Everyone needs time to recharge. Eight hours of sleep and two hours of break each day still leaves thirteen hours a day to do important stuff, and I certainly have not optimized that yet.

Prep -- I’m glad that living closer to work has finally brought this number down to something I feel is manageable.

Planning -- I spent significantly more time planning Oct-Dec ‘14 than I did in Jul-Sep ‘14. I don’t know why that is.

Email -- I’m spending more time on my email now. I think this is because I do get more email, but I think it’s more that I’m spending more concentrated periods of time doing emails and multitask less while handling email, so I’m more likely to tag time as “Email”.

Cleaning -- I’m spending a lot more time cleaning my room now in Oct-Dec ‘14 than Jul-Sep ‘14. It doesn’t feel that way, though. I wonder why that is? I imagine that maybe my Jul-Sep ‘14 number wasn’t very accurate?

Moving -- I’m delighted that after spending a ton of time moving in June and September, I finally have a review period where I spent no time moving!

 

Conclusion

I find it very hard to make resolutions for the entire year -- I just don’t know enough about what my life will be like to plan that far. I much rather like making goals for the next three months instead -- it’s long enough to be a serious commitment, but short enough to allow you to pivot.

I believe that while Oct-Dec ‘14 didn’t bring me as many accomplishments as I hope to have, it did set up a lot in the background that will make Jan-Mar ‘15 unusually productive. I’m excited to work more on Charity Science and I’m excited to get some entrepreneurship projects off the ground and hopefully start earning some money from places other than my job.

However, while I’m reflecting, I must note that I talk a lot a lot about wanting to do more things, but I don’t talk enough about what I’m going to cut out to make that happen. I’ve long run out of the delusion that I can just make more time for everything, but I don’t always act on it.

Ideally, I think my week should have five more hours spent on my day job, five more hours spent on Charity Science, five more hours spent on Entrepreneurship projects, and 30 more minutes spent on learning Chinese. To make way for this, I’m going to reap the savings of finishing the EA Survey (-4.4hrs/wk) and am going to try to dial back socializing by 11.1hrs/wk.

Update -- I don't think I have fine-grained control over my schedule, but I'd like to aim for:

-10.6hrs/wk Social (to 18.6hrs/wk)
-4.4 EA Survey (to 0)
-1.4 Birthday for Charity (to 0)
+3.3 Charity Science (to 5)
+2.6 Exercise (to 4)
+0.5 Chinese (to 2)
+2.2 Learn Programming/Stats/Math (to 4) (see approach here)
+2.5 Day Job (to 28)
+5.3 Entreprenuership (to 8)

...We’ll see how that goes.

 

What you can do to help!

Let me know if (a) this post is useful to you or (b) you'd rather I not post here on the EA forum, or somewhere in between ...or I guess you can hold both (a) and (b) consistently, so it's not an either-or.

Also let me know if there's anything I can do to make this review more useful, or if you want to know something I didn't provide!

Let me know if you think I should stop doing something I'm currently doing.

Let me know if there's something I'm not currently doing that I should be doing.

Comments (20)

Comment author: Pablo_Stafforini 03 January 2015 06:49:38AM *  6 points [-]

Let me know if (a) this post is useful to you or (b) you'd rather I not post here on the EA forum

I found this post useful and hope you continue to post your personal reviews in this forum.

Comment author: RyanCarey 03 January 2015 12:09:55PM 0 points [-]

Yeah I think it's nice to see people do more of it, even if they're shorter or less systematic updates.

Comment author: Bitton 03 January 2015 04:52:00PM 2 points [-]

For the record, I'm 2 weeks into my first Peter-style personal review.

Comment author: Peter_Hurford  (EA Profile) 03 January 2015 06:35:54PM 0 points [-]

I'm excited to read it!

Comment author: Peter_Hurford  (EA Profile) 03 January 2015 06:37:02PM 1 point [-]

I think it would be cool to get an update from you, Ryan. I frequently think to myself "Ryan has tried a bunch of things between earning to give and working for CEA, and now I'm not really sure what he's up to, but if he would just outline his thoughts on that, that would be really neat!" And then I think to myself that I should just ask you to Skype, but I forget.

I also do the same thing with Ben Kuhn.

Comment author: RyanCarey 03 January 2015 06:50:15PM *  1 point [-]

Yeah, I also wonder about Ben from time to time.

My short version is that I nominally spent a quarter of 2015 on each of medical work, being at Leverage, being at 80k and being a student in Imperial's M Sc of Bioinformatics.

Voluntarily, I've run this forum and helped CSER.

Of those, I think helping CSER has gone best, while medical work, MSc and 80k have also gone ok.

My favourite achievement of the year has been studying machine learning. The bottom line is that I'm unclear where the best place for me to be is on a scale between researching medical imaging algorithms academically, making some kind of health AI company, working in research assistance/management for CSER or earning to give in medicine and I expect to have a better idea by the time I finish the degree in 9 months.

Comment author: Ben_Kuhn 04 January 2015 08:07:33PM 1 point [-]

Huh. I haven't done these because I figured everything I've done recently has been available to e.g. people who read my blog. (Also because I don't use Toggl like Peter does so they would be way less exciting/factual.) What kind of stuff would you (or Peter) want to see?

PS While we're on the subject, you should blog more!

Comment author: RyanCarey 04 January 2015 08:25:27PM 1 point [-]

I'd be interested in what you're doing, which skills you're using, how it compares to some plausible alternatives in interestingness/usefulness/likely progression, and any valuable tips you'd give to someone similar to you e.g. graduating from math that weren't obvious at the time.

I'll likely blog more when some perspective seems neglected by posters here or when I have a new project to plan and promote.

Comment author: Peter_Hurford  (EA Profile) 05 January 2015 04:53:30AM 0 points [-]

Agree with Ryan here, Ben. I'd love to hear more about Theorem LP, your future plans, path to impact, any EA side projects, etc.

Comment author: Peter_Hurford  (EA Profile) 03 January 2015 11:25:15PM 0 points [-]

Awesome, thanks! I look forward to further updates. And we definitely should talk more about machine learning (and start-ups) sometime.

Comment author: kdbscott 03 January 2015 05:35:56PM 2 points [-]

Thank you for posting this Peter, I find these useful!

A couple questions:

1) How do you use Toggl? Desktop app? Mobile? Do you also use Rescuetime? I've found it difficult to use Toggl when I'm away from the computer, during social/break/food/cleaning/interruptions/errands/other time. Maybe I just need to get in the habit of pulling out my phone and start/stopping the timer? 2) How do you do food? I spend a lot of time on food (buying, cooking, eating) each week, and I enjoy it but would be open to time saving techniques (especially when it comes to veggies)

Thanks again!

Comment author: Peter_Hurford  (EA Profile) 03 January 2015 06:49:50PM 1 point [-]

How do you use Toggl? Desktop app? Mobile?

I use the mobile app a lot, because frequently I need to update while I'm on the go.

While on my computer, I actually use the command line interface plus a custom wrapper I wrote to make the CLI more user-friendly and more customized to how I personally use Toggl. Before I used that, though, I used the web interface. I was unaware there was a desktop app.

-

Do you also use Rescuetime?

No, I've never used Rescuetime. The core problem I have is that sometime me using Facebook is goofing off and sometimes me using Facebook is for important work. Rescuetime can't tell the difference, but I can.

-

I've found it difficult to use Toggl when I'm away from the computer, during social/break/food/cleaning/interruptions/errands/other time. Maybe I just need to get in the habit of pulling out my phone and start/stopping the timer?

That's my solution.

-

2) How do you do food?

Of note is that "buying food" is categorized as "Errands" and "eating food" is usually a part of "Break", so my total time spent on food-related activities is higher than just the "Food" category (which is solely for assembling the food) would imply. Maybe I should fix that if it would be useful to include a more broad assessment of food (buying, assembling, AND eating)?

-

I spend a lot of time on food (buying, cooking, eating) each week, and I enjoy it but would be open to time saving techniques (especially when it comes to veggies)

This has been the most requested (and by that, I mean from four people) feature, so I'll probably write a post on this, but I think it would disappoint people by being too specific to my personal circumstances (and by that, I mean weirdness).

Comment author: kdbscott 03 January 2015 07:40:51PM 0 points [-]

I was unaware there was a desktop app. There is, and they just released global shortcuts, which has helped a lot!

I'll probably write a post on this, but I think it would disappoint people If it's too weird I'll just stop reading! I like my familiarity points system, but when you publish everything online it might be prudent to consider weirdness. Perhaps a shareable document for those interested?

Comment author: RyanCarey 03 January 2015 02:08:30AM *  0 points [-]

Looks great, Peter. Annual reviews are pretty personal. Do you want suggestions of things to do/not to do in your next year? Ways to make the review better? Probing questions, et cetera?

Comment author: Peter_Hurford  (EA Profile) 03 January 2015 02:30:22AM 0 points [-]

Thanks -- added a final section on that:

Let me know if (a) this post is useful to you or (b) you'd rather I not post here on the EA forum, or somewhere in between. Also let me know if there's anything I can do to make this review more useful, or if you want to know something I didn't provide!

Let me know if you think I should stop doing something I'm currently doing.

Let me know if there's something I'm not currently doing that I should be doing.

Comment author: RyanCarey 03 January 2015 12:38:30PM 0 points [-]

Sweet. My thoughts:

48 hours a week on job/ea/planning/email/entrepreneurship/programming is pretty good. EAs that I know have expressed that they're very impressed with how much you seem to get done, so you can be proud of that.

I think you're going to have trouble being a successful entrepreneur using only three hours, unless you're somehow able to outsource all of the jobs effectively. You still might find 'dipping your toe into the waters' to be valuable though.

I think this is smart, because the only way I can reliably know whether they’ll work is to actually do them -- I’m not really able to tell in advance.

Makes sense. Although considerations of cause prioritisation can help with part of the question. To me, guiding tech development in light of risks seems important. Animal welfare lobbying not so much. Building infrastructure for EA projects - quite unclear. So the Survey/Blogging/.impact might be useful if you think building this infrastructure is useful, although 6.4 hours per week is a lot. Helping charity science looks similar - potentially decently useful.

One thing I did this year that I expect to pay large dividends was that I completed Andrew Ng's Machine Learning course on Coursera. I would recommend this (or if you have already, then proceed to complete the more detailed Stanford Course notes) to reinforce your stats knowledge, help you apply for higher paying jobs and perhaps give some insights into next steps for getting involved with guiding tech development.

Apart from thinking about cause prioritisation and the ML course, my other question is - have you considered which location is optimal for you to be in? SF is good to visit or live. Presently, two of my Melbourne friends who are EAs have moved there and three more are in the process of doing so. Brayden, Chris, Frazer, Tara and Helen. It has unsurpassed EA connections (GW, tech entrepreneurship, LW, MIRI, CFAR, Thiel, Leverage), higher pay, progressive overall culture and good social groups. Obviously, where to live (or visit) is very person-specific but from the outside, it looks valuable to go there to visit or live.

Hope those thoughts help!

Comment author: Peter_Hurford  (EA Profile) 03 January 2015 07:20:27PM 1 point [-]

48 hours a week on job/ea/planning/email/entrepreneurship/programming is pretty good. EAs that I know have expressed that they're very impressed with how much you seem to get done, so you can be proud of that.

Thanks!

-

I think you're going to have trouble being a successful entrepreneur using only three hours, unless you're somehow able to outsource all of the jobs effectively. You still might find 'dipping your toe into the waters' to be valuable though.

I agree on both parts of this. It doesn't take too much time to MVP a product, but I would like to ramp up my time here. (I'm proposing to myself to aim for eight hours a week, though I don't have super-fine-grained schedule control.) I think the low time investment is part of the reason why we didn't launch anything.

-

considerations of cause prioritisation can help with part of the question. To me, guiding tech development in light of risks seems important. Animal welfare lobbying not so much.

I agree with cause prioritization being important, and definitely owe more thought to the importance of guiding tech development and what role I can personally play in that.

As to animal welfare lobbying, I think my interest will rise or fall quite dramatically based on the outcome of the veg study. I personally think that launching that study will be a quite powerful contribution to cause prioritization, which was my goal.

-

Building infrastructure for EA projects - quite unclear. So the Survey/Blogging/.impact might be useful if you think building this infrastructure is useful, although 6.4 hours per week is a lot.

I agree that I've hit diminishing marginal returns here and these things are not as high-impact as other things I can be doing and I aim to try to walk away from them.

-

One thing I did this year that I expect to pay large dividends was that I completed Andrew Ng's Machine Learning course on Coursera. I would recommend this (or if you have already, then proceed to complete the more detailed Stanford Course notes) to reinforce your stats knowledge, help you apply for higher paying jobs and perhaps give some insights into next steps for getting involved with guiding tech development.

This plays in quite well with some educational goals that I already have with regard to my day job that I've started recently. My three point plan is as follows:

Boost machine learning knowledge with Introduction to Statistical Learning and the associated Stanford online course. This is what my boss at work recommended to me. I'm not sure if it's better than Andrew NG's course.

Boost R knowledge with Advanced R. I didn't know R well enough the first time around to fully appreciate this book, but now I think I do.

Boost math knowledge with Khan Academy. I wanted to learn linear algebra because it helps with machine learning, but in order to learn linear algebra I needed to brush up on my Calculus II. Well, my Calculus II wasn't that good, so I need to brush up on my Calculus I. Calculus I also didn't go so well, so now I'm at Algebra II. ...I'll have to level up my math knowledge from there.

...Currently I'd aim to spend two hours a week on that, but you're right I might want to spend more. Maybe instead of increasing in my day job as much, I can spend some more time on this.

-

Apart from thinking about cause prioritisation and the ML course, my other question is - have you considered which location is optimal for you to be in? SF is good to visit or live.

I'd love to visit SF for sure (I already meant to a few times, but never got around to it) and I would not be too surprised if I worked there at some point in the future. Right now, the job I have is in Chicago, though, and I think the benefits outweigh the costs for leaving my job at the moment, and I foresee that being true for at least the next six months.

-

Hope those thoughts help!

Sure did! Thank you for your time!

Comment author: RyanCarey 03 January 2015 08:46:38PM 1 point [-]

Great, it seems like our views are pretty harmonised.

Boost machine learning knowledge with Introduction to Statistical Learning and the associated Stanford online course. This is what my boss at work recommended to me. I'm not sure if it's better than Andrew NG's course.

Yeah, I read Introduction to Statistical Learning in R (ISLR) and then went on to Ng's course. It was when I viewed the latter that it really clicked. I still think ISLR was good though. Ng uses Matlab (bad if you're getting used to R), provides better explanations for the math (good), has quizzes (good) and focuses more on modern approaches including neural networks (good). So I would suggest at least enrolling in the Coursera course in order to have it available as a backup.

Re math prerequisites, Khan Academy is also how I got to being able to understand machine learning. I also did a Coursera course in Calculus, which might help. This Jim Fowler is a good teacher, similarly to Sal Khan. To start learning machine learning, you'll need basic linear algebra - how to multiply vectors and matrices basically. For the introductory stuff, you can apply it without understanding calculus, although it's fundamental to how machine learning works and you'll need to learn some calculus sooner or later. But if it helps with motivation, if you know the basics of how to multiply vectors and matrices, you can definitely start Ng's course, and learn some of the rest in tandem. In the long-run, to get good at ML clearly requires proficiency with linear algebra, calculus and statistics, as well as a general comfort with math that seems to come from practice, so it seems like we have to put in a bunch of solid hours into getting these foundations. More hours per week seems good.

I'd love to visit SF for sure (I already meant to a few times, but never got around to it) and I would not be too surprised if I worked there at some point in the future. Right now, the job I have is in Chicago, though, and I think the benefits outweigh the costs for leaving my job at the moment, and I foresee that being true for at least the next six months.

I think that the several hundred dollars it costs you to make a flight there, and a week of lost salary would pay itself off in expected future earnings from the option of working there later, new professional contacts, new insights into cause prioritisation, fun, new friends, extra passion for doing good, etc. The new Global EA Summit will be coming up too, which might be a reason to get down there. You meeting people there just feels like all-round good news to me.

Comment author: Peter_Hurford  (EA Profile) 03 January 2015 11:34:02PM *  1 point [-]

Yeah, I read Introduction to Statistical Learning in R (ISLR) and then went on to Ng's course. It was when I viewed the latter that it really clicked.

Makes sense. I think I'll try that in the same order as well.

Re math prerequisites, Khan Academy is also how I got to being able to understand machine learning. I also did a Coursera course in Calculus, which might help.

I don't actually know the basics of multiplying vectors and matricies (I learned them in college but forgot soon afterward), so I should learn that first.

You've convinced me of two changes to make:

  • First, I should go in sequence with my learning rather than parallel. I think I'll aim for Khan Academy Algebra -> Khan Academy Calculus I -> Khan Academy Calculus II -> Khan Academy Linear Algebra -> Introduction to Statistical Learning -> Angrew Ng's course. (I think I'll still do Advanced R --> Learn Hadoop in parallel, though, because my R skills are somewhat unrelated to my MR skills.) (To-do for self: re-arrange learning list.)

  • Second, I should spend more than 2hrs/wk on this. I can probably cut out more EA time. (To-do for self: think on this more.)

-

The new Global EA Summit will be coming up too, which might be a reason to get down there. You meeting people there just feels like all-round good news to me.

Yeah, I'll come out for the Global EA Summit and you've convinced me to try to make a full week of it. We have a pretty flexible vacation policy here, so I shouldn't even lose salary. I just have to reconcile this with other vacation I plan on taking. (To-do for self: plan out vacation for 2015, watch for SF EA Summit dates.)

I'm thinking of also going to SF when Joey and Xio get around to visiting SF, but I don't know if that's going to be in 2015.

Comment author: RyanCarey 04 January 2015 12:42:13PM 0 points [-]

Excellent,

Good luck!