最新の記事
4,485 件並び順
20:33EN
3Blue1Brown· 2017年10月16日Gradient Descent: How a Neural Network LearnsGrant Sanderson, Leisha Lee · 3Blue1BrownA visual introduction to minimizing training cost, and a look at what a digit classifier does and does not learn.
18:18EN3Blue1Brown· 2017年7月21日Hilbert’s Curve: What an Infinite Path Reveals About Finite ImagesA recursive path through pixels leads to a continuous curve that fills a square.
1:38:23EN
Jane Street· 2017年7月18日Yudkowsky and Hanson Debate Whether an Intelligence Explosion Would Be LocalEliezer Yudkowsky, Robin HansonA historical debate about self-improving AI, accumulated knowledge, innovation and whether a first mover could gain a decisive lead.
52:51ENJane Street· 2017年6月20日Caveat Configurator: When Configuration Becomes CodeDominick LoBraico · Jane StreetDominick LoBraico explains why flexible OCaml plugins created production problems at Jane Street, and how configuration generation and disciplined plugin management address them.
5:38EN3Blue1Brown· 2017年5月7日Higher-Order Derivatives: Curvature, Acceleration and JerkGrant SandersonA visual introduction to derivatives of derivatives.
22:19EN3Blue1Brown· 2017年5月7日Taylor Series: Turning Derivatives Into ApproximationsBuild a cosine approximation, see the quadratic term geometrically, and understand convergence.
12:39EN3Blue1Brown· 2017年5月6日Why Average Height and Average Slope Connect Integrals and Derivatives3Blue1Brown uses the average of sin x to derive an intuitive link between area and the slope of an antiderivative.
16:50EN3Blue1Brown· 2017年4月29日The Derivative and the Paradox of Instantaneous ChangeGrant SandersonA geometric introduction to difference quotients, limits and the derivative of a cubic.
59:14ENJane Street· 2017年3月15日Arjun Guha on Verifying System ConfigurationsArjun Guha · UMass AmherstPuppet manifests can be declarative yet nondeterministic or non-idempotent. A verification tool models resource interactions to find those problems.
38:17ENJane Street· 2017年3月13日How to Build a Reliable Exchange with Deterministic ReplayBrian Nigito · Jane StreetA matching engine, multicast and state-machine replication can turn a high-throughput exchange into a recoverable system. This talk explains the design and its tradeoffs.
25:44ENJane Street· 2017年3月9日Why Functional Programming Matters at Jane StreetSebastian Funk · Jane StreetA Jane Street developer revisits John Hughes’s classic argument and explains why expressive static types rank highest for trading systems.
28:31ENJane Street· 2017年3月9日Incremental: Self-Adjusting Computations in OCamlSebastian Funk · Jane StreetSebastian explains Jane Street’s dynamic dependency graph library and the algorithms behind efficient updates.
13:58EN3Blue1Brown· 2016年11月25日Solving the Towers of Hanoi by Counting in BinaryKeith SchwarzBinary carries select the disk to move, and a recursive argument explains why the solution uses the fewest possible moves.
16:29EN3Blue1Brown· 2016年11月4日Why Topology Matters: An Inscribed Rectangle ArgumentGrant SandersonGrant Sanderson turns pairs of points on a curve into a Möbius strip.
1:29:03EN
Lex Fridman· 2016年9月27日ポッドキャストFrom Word Vectors to Memory Networks: The Building Blocks of Deep NLPホスト Lex FridmanRichard Socher explains how word embeddings, recurrent models, GRUs, pointer mixtures, and dynamic memory networks turn language tasks into trainable end-to-end systems.
13:10EN3Blue1Brown· 2016年9月1日Why Cross-Product Coordinates Match Their GeometryGrant SandersonA linear transformation and duality connect the determinant formula to area and direction.
43:16EN
GOTO Conferences· 2016年8月29日How Platforms Work: From Deployment to RecoveryCasey WestCasey West outlines six platform capabilities and demonstrates them with Cloud Foundry in 2016.
30:03EN
Google DeepMind· 2016年7月5日Mustafa Suleyman’s 2016 Vision for Technology in Patient CareMustafa Suleyman · DeepMindA historical talk on clinician-led design, kidney-injury alerts, eye-scan research and the need for scrutiny.
1:09:24ENJane Street· 2016年4月5日Yaron Minsky on Effective ML: Writing OCaml for Code ReadersYaron Minsky · Jane StreetPractical lessons from Jane Street on using types, interfaces and abstraction to keep a large OCaml codebase understandable and correct.
16:01EN3Blue1Brown· 2016年4月1日The Brachistochrone: Why the Fastest Descent Is a CycloidGrant Sanderson, Steven StrogatzGrant Sanderson and Steven Strogatz explore Bernoulli’s light analogy and a geometric proof.
1:06:39ENJane Street· 2016年3月31日Why Jane Street Uses OCamlYaron Minsky · Jane StreetA Jane Street talk explains how functional programming and static types support reliable trading software, and examines the costs of choosing a less common language.
23:58ENJane Street· 2016年3月31日Effective ML: Errors, Abstraction and Managing ComplexityYaron Minsky · Jane StreetA Jane Street engineer discusses explicit failure, boilerplate, type-system restraint and the practical role of side effects.
1:04:56ENJane Street· 2016年3月31日Yaron Minsky on Incremental: From Research Idea to Production LibraryYaron Minsky · Jane StreetA history of the algorithms, memory problems and performance tradeoffs behind Jane Street’s OCaml library for self-adjusting computations.
1:12:41ENJane Street· 2016年3月31日Ron Minsky on Trading, Correctness and Jane Street’s Shift to OCamlYaron Minsky · Jane StreetA Jane Street technologist explains the market-making business and why readability, types and predictable performance shaped the firm’s choice of OCaml.もっと見る4,485 件中 24 件を表示