loehnsberg

10 ms·

61.
by loehnsberg 3y ago
From my own experience, a properly cross-validated lasso regression over a wide range of autoregressive features beats FB Prophet by a good margin and offers nearly the same degree of automation.
62.
by loehnsberg 3y ago
Based on your searches, the sites that you visit, the videos that you watch Google or the likes create a profile that they then offers to companies to target you with ads. The effect may be subtle, but ads work, and companies hand millions
63.
by loehnsberg 3y ago
The thing is you ARE paying for search but do not know how much, because you will never know where a targeted ad nudged you to pay a little markup for something that you bought. I prefer to pay for the services that I use with money rather
64.
by loehnsberg 3y ago
The iPhone mini is 135g. You can use it with one hand. It fits into any pocket. Take it with you when running (or swimming) and there‘s no need for an Apple Watch. Also, you don‘t need to twist your hand when paying but just pull it out of
65.
by loehnsberg 3y ago
A hyperplane is a multi-dimensional linear function that splits space into two distinct regions. In the context of a classifier, it splits feature space into disjunct sub-spaces (one for each class). SVMs effectively place a hyperplane with
66.
by loehnsberg 3y ago
The file manager has gotten better but it is still way behind Finder. Otherwise, pair it with a keyboard and mouse, open codespaces, and you have an excellent travel computer. Just don‘t expect to move files around.
67.
by loehnsberg 3y ago
The buckets are 1,…,100, with percentiles being their boundaries. Since the 0th and 100th percentile are insufficient to form an interval, we only practically use percentiles 1,…,99.
68.
by loehnsberg 3y ago
Germany has a lot of things going for it, but it‘s weaknesses become really apparent at the moment: - finance happens elsewhere, no world-leading banks anymore, meaning it‘s difficult to get capital, unless you‘re big corp - no natural ress
69.
by loehnsberg 3y ago
To be fair, video calls on Slack were below par at the time when Teams came out, and little effort was made by Slack to improve the call feature in the months that followed.
70.
by loehnsberg 3y ago
Thanks for sharing. That‘s exactly what I was looking for.
71.
by loehnsberg 3y ago
ISL is the best intro-level textbook of classic ML methods. It‘s theory-oriented yet simple enough to appeal to a wide audience of students (with basic knowledge in stats, linear algebra, and coding). Having the examples only in R was a pai
72.
by loehnsberg 3y ago
I use an iPad Pro mostly for my lectures and for reading and annotating PDFs. After the latest upgrade of iPadOS, I also explored it for more serious work connecting it to a mouse and keyboard. If you're happy with working inside a bro
73.
by loehnsberg 3y ago
Would be nice to play this on iOS. Just probably not worth the hassle for the relatively small number of iPad M1/M2 owners.
74.
by loehnsberg 3y ago
Couldn‘t the fully connected 3x3x3 network in Figure 2c simply be reformulated as an equivalent 9x9 network that is not fully connected? Instead connections between layers of the sub-networks would only occur every 3 layers?
75.
by loehnsberg 3y ago
Take a look at Kagi Search [1] where you pay with money instead of your data. Have been using it for over a year now and must say that the search engine is surprisingly good. The only time I ever use !g bangs is when I want live updates of
76.
by loehnsberg 4y ago
I have been testing it extensibly in the past weeks. It is a great tool for writing. I had it solve a data science exam that I give to my students. It passed with flying banners. It even got close in the parts that involved calculations and
77.
by loehnsberg 4y ago
+1 for Kagi: I have been using it for a year now. I still use Google on my phone (Kagi's own browser Orion is still a bit buggy), so I can compare daily. Classic search results are comparable, with lenses they are even better, plus Kag
78.
by loehnsberg 4y ago
I used the dataset with my students, as it is small and does not require preprocessing, like dummy coding or handling missing values. Students also brought the racial issue to my attention and it created a bit of a discussion. We eventually
79.
by loehnsberg 4y ago
If you work for your e-company remotely as an e-resident from another EU country, you still need to get a trade license and VAT number to write invoices to your e-company and file income tax. If we leave liability aspects aside for a moment
80.
by loehnsberg 4y ago
The potential for hydro-electricity is mostly exploited, and it does not offer a solution for the long-term storage of electricity for large parts of the world. Not all countries are like Norway (or Brazil or Canada). There is still space l
81.
by loehnsberg 4y ago
M/M/s queues have nice properties, but require that arrival and service times are exponentially distributed. Once you deviate from this assumption, closed-form solutions become difficult to obtain and numerical approximations are
82.
by loehnsberg 4y ago
Employees from the HN sphere maybe, but ask a temp worker. They can get fired on the spot — in Germany!
83.
by loehnsberg 5y ago
Score cards are simple but prone to manipulation. Another approach from multi-criteria decision-making that is very useful for benchmarking is Data Envelopment Analysis (DEA) [0], or the Analytic Hierarchy Process (AHP) [1] for group-decisi
84.
by loehnsberg 5y ago
The following about the sentiment towards nuclear power in Germany: - Germany is a densely populated country and Germans have always been sensitive towards environmental pollution. - Nuclear fallout is a reality as every consumer of mushro
85.
by loehnsberg 5y ago
I can only agree. Used it for a couple of weeks when live co-coding was indispensible. You cannot even compare this to similar features in Teams, Slack, and the like. The tool works seamlessly and has a super-smooth usability. I did not not
86.
by loehnsberg 5y ago
Sapphire glass has also been used in premium wrist watches for ages. These watches don't scratch. And when you drop the watch on the face from table height, say, it's more likely that the mechanics inside break than the glass. I s
87.
by loehnsberg 5y ago
I often have really good ideas under the shower that eluded me while thinking hard about it at my desk. I really wonder when they can reproduce the creative boost of taking a shower in a lab experiment.
88.
by loehnsberg 5y ago
I am not an expert on convex, unconstrained optimization but I frequently use these methods. I'd wait until it's peer-reviewed and published in a good journal (SIAM) before I get too excited about this one.
89.
by loehnsberg 5y ago
Once iOS and MacOS are merged, will Apple open up its app store to the wild west or will every little OS hack have to depend on their approval? I would certainly be put off by the latter. I was willing to accept the walled garden for my lit
90.
by loehnsberg 5y ago
Thanks for sharing all of these examples. It is great to see that you are using Anylogic which I think is an excellent workbench for creating discrete-event models of production systems and entire supply chains even. Anylogic also comes wit

More ›