Inception


README.md

This repository contains a reference pre-trained network for the Inception
model, complementing the Google publication

Going Deeper with Convolutions, CVPR 2015.
Christian Szegedy, Wei Liu, Yangqing Jia, Pierre Sermanet, Scott Reed,
Dragomir Anguelov, Dumitru Erhan, Vincent Vanhoucke, Andrew Rabinovich

You can view “inception.ipynb” directly on github, or clone the
repository, install dependencies listed in the notebook and play with code
locally.

You may also be intersted in the Multibox
approach that uses the Inception architecture for object detection, also
available on github.

Disclaimer: this is not an official Google product (experimental or otherwise).


Original URL: http://feedproxy.google.com/~r/feedsapi/BwPx/~3/6hG-NjEQbX4/inception

Original article

Department of Education Releases 20 Years of College Earnings Data

While there is variation in the amount of debt and fraction of students borrowing by sector, on average, students at private for-profit two-year and four-year institutions have high rates of borrowing and their graduates often have large amounts of debt. While debt per se may not be problematic where students are able to repay their loans, it should be paired with other data, such as completion rates and post-school earnings, to provide a more comprehensive picture of student outcomes.

Learn more in the Technical Paper

Loan repayment rates offer a new way to consider borrower behavior, and may convey more information about student debt management than cohort default rates (CDR). The overall three-year repayment rate for all undergraduate institutions, weighted by the number of students borrowing at each school, was 63 percent in the combined 2010 and 2011 repayment cohorts. 37 percent of students were thus not meeting the repayment metric—either they were in default or were making monthly payments that were not reducing their loan balance. As a point of comparison, the three-year CDR was 13 percent for all students in the 2011 repayment cohort.

Learn more in the Technical Paper

There is a great deal of earnings variation within a college, so that students with good outcomes at low-earning schools often perform better than students with poor outcomes at high-earning schools. Earnings variation can be explained at least in part by the program in which students choose to study; program-level earnings data will be available moving forward, beginning with the 2012 cohort.

Learn more in the Technical Paper

The data in the National Student Loan Data System (NSLDS) on enrollment intensity and transfer status are both of poor quality for Pell-only students prior to 2012. Because of this, the data do not support reporting completion rates disaggregated by full-time and part-time status, or first-time and not-first-time status. Moreover, since transfers can only be identified if the student receives Title IV aid at the transfer-in institution, NSLDS cannot reliably identify all transfer students.

Learn more in the Technical Paper

To help complement federal data and address the limitations of the federal graduation rate measure collected through IPEDS, several external organizations have launched efforts to measure accurate completion rates and other data sources. One effort, the Student Achievement Measure, provides data on progress and completion of transfer, part-time, and full-time students. The Voluntary Framework of Accountability explores remedial education, academic measures, and workforce outcomes. Additionally, IPEDS will begin collecting and publishing more comprehensive graduation rates next year.

Learn more in the Technical Paper


Original URL: http://feedproxy.google.com/~r/feedsapi/BwPx/~3/IrT3Sr16wYU/

Original article

Experiential Learning: ABA Standards 303 and 304

Although “experiential learning” has been a term of interest to many legal educators for years, the ABA’s new standards have brought it front and center by mandating that schools “require[] each student to satisfactorily complete at least . . . one or more experiential course(s) totaling at least six credit hours.” Two well-known ways to meet the requirement are […]


Original URL: http://bestpracticeslegaled.albanylawblogs.org/2015/09/13/experiential-learning-aba-standards-303-and-304/

Original article

What is Ethereum, 2nd gen. cryptocurrency, good for? – a visual example


Hi, I’m a pyramid scheme contract living on the Ethereum blockchain.

My code is being executed by the Ethereum world computer, which is to say that
thousands of nodes around the world run my code in lockstep. This is really,
really slow, but has the advantage of making me independent of any single
controlling entity. Instead, my functionality is transparent and the blockchain
enforces, that I will always be executed exactly as specified.

You can take a look at my source code, but the gist
is, that

by sending 1 ETH to me, you are entered into the pyramid

and stand to more than double your money if the level below you fills up
(triple, minus a 10 % fee, to be exact). Ethereum contracts sleep most of the
time and only wake up and run when a transaction is sent to them. Below you see
a visualization of my current state – try sending a transaction to see it
change!

To be clear: This is less about getting insanely rich and more about
demonstrating what an Ethereum contract can do. Where previously you would
have needed a middleman, who has a lot of control over the process
(maintains a list of participants, administers payouts, etc.), you can now have
a programmable, transparent and incorruptible middleman
on the blockchain.

For a second point of view, try looking me up on any
Ethereum blockchain explorer
– like EtherCamp,
Etherscan
or Etherchain.


Original URL: http://feedproxy.google.com/~r/feedsapi/BwPx/~3/73CIq7c5UJY/

Original article

Proudly powered by WordPress | Theme: Baskerville 2 by Anders Noren.

Up ↑

%d bloggers like this: