It’s winter, it’s the holidays and it’s quarantine-times: It’s the perfect recipe for doing a ton of baking. In fact, U.S. search interest in "baking" spiked in both November and December 2020.
But being in the AI field, we decided to dive a little deeper into the trend and
try to understand the science behind what makes cookies crunchy, cake spongy and bread fluffy — and we decided to do it with the help of machine learning. Plus, we used our ML model to come up with two completely new baking recipes: a cakie (cake-cookie hybrid) and a breakie (bread-cookie hybrid). (Don’t worry, recipes included below.)
We started off by collecting hundreds of cookie, cake and bread recipes. Then we converted all of their ingredients to ounces and whittled them down to a few essential ingredients (yeast, flour, sugar, eggs, butter and a few other things). Next we did a bit of reorganizing, since according to Paul Hollywood, treats like banana, zucchini and pumpkin bread are really more cake than they are bread.
Then we used a Google Cloud tool called AutoML Tables to build a machine learning model that analyzed a recipe’s ingredient amounts and predicted whether it was a recipe for cookies, cake or bread. If you’ve never tried AutoML Tables, it’s a code-free way to build models from the type of data you’d find in a spreadsheet like numbers and categories – no data science background required.
Our model was able to accurately tag breads, cookies and cakes, but could also identify recipes it deemed “hybrids” — something that’s, say, 50% cake and 50% bread, or something that’s 50% cake and 50% cookie. We named two such combinations the “breakie” (a bread-cookie — "brookie” was already taken) and the “cakie” (a cake-cookie) respectively.
Being science-minded bakers, we had to experimentally verify if these hybrid treats could really be made. You know, for science.
Behold the cakie: It has the crispiness of a cookie and the, well, “cakiness” of a cake.
We also made breakies, which were more like fluffy cookies, almost the consistency of a muffin.
Sara's first batch of breakies.
Beyond just generating recipes, we also used our model to understand what made the consistency of cookies, cakes and breads so different. For that, we used a metric called “feature importance,” which is automatically calculated by AutoML Tables.
In our case, the amount of butter, sugar, yeast and egg in a recipe all seemed to be important indicators of “cookieness” (or cakiness or breadiness). AutoML Tables lets you look at feature importance both for your model as a whole and for individual predictions. Below are the most important features for our model as a whole, meaning these ingredients were the biggest signals for our model across many different cake, cookie and bread recipes:
If you find yourself with extra time and an experimental spirit, try out our recipes and let us know what you think. And you can find all the details of what we learned from our ML model in the technical blog post.
Most importantly, if you come up with an even better cakie or breakie recipe, please let us know.
https://j.mp/3rWkXOq
WARNING: Cookies may contain information for a customized Web page or logon information for a Web site. Before you delete your cookies, you may want to export or save them. If you are using Internet Explorer 8.0 or later, use the Import/Export Wizard in Internet Explorer to export your cookies.
Tuesday, January 5, 2021
Most Reliable Hosting Company Sites in December 2020
Rank
Performance Graph
OS
Outage
hh:mm:ss Failed
Req% DNS
Connect
First
byte Total
1
www.choopa.com
Linux
0:00:00
0.000
0.270
0.005
0.024
0.024
2
Webair
Linux
0:00:00
0.000
0.327
0.071
0.141
0.141
3
krystal.uk
Linux
0:00:00
0.000
0.255
0.088
0.176
0.176
4
GoDaddy.com Inc
Linux
0:00:00
0.005
0.409
0.009
0.042
0.044
5
Rackspace
Linux
0:00:00
0.005
0.466
0.010
0.021
0.021
6
New York Internet (NYI)
FreeBSD
0:00:00
0.005
0.556
0.058
0.116
0.116
7
Hyve Managed Hosting
Linux
0:00:00
0.005
0.181
0.077
0.153
0.153
8
CWCS Managed Hosting
Linux
0:00:00
0.005
0.338
0.082
0.166
0.166
9
ServerStack
Linux
0:00:00
0.005
0.251
0.083
0.166
0.166
10
www.dinahosting.com
Linux
0:00:00
0.010
0.291
0.094
0.189
0.189
See full table
Choopa.com finished 2020 as the most reliable hosting company site in December, with no failed requests and the fastest average connection time. Choopa.com topped the table four times in 2020, more than any other hosting company site, and made a total of eight appearances in the top 10. The hosting provider offers cloud hosting, dedicated hosting and colocation in its own primary facility in Piscataway, New Jersey as well as smaller facilities in Los Angeles, Amsterdam, and Tokyo.
https://j.mp/3pUuLqi
https://j.mp/3pUuLqi
Jumpstarting your January with key product updates
Welcome everyone to 2021! We’re now past the holidays and looking towards what will hopefully turn out to be a much better year than the last (certainly a low bar to clear!) As we put 2020 in the rearview mirror and look ahead, there’s lots to get excited about, from a public health standpoint but also to help further empower you in the digital advertising space. At Microsoft Advertising, our goal is for you to reach the right audience with the right message and the right technology, and here are this month’s key product updates to help you towards that goal.
This month’s top story: Platform-wide brand safety protection now expanded to Canada and the UK
Back in September, we announced our new US partnership with Integral Ad Science (IAS), the global leader in digital ad verification, to help advertisers achieve their advertising goals while implementing the right guardrails to protect their brand. We’re delighted to share that this partnership has now expanded to Canada and the United Kingdom, offering an additional layer of brand safety protection on the Microsoft Audience Network in those markets at no additional cost.
Learn more about the partnership in the September post, and stay tuned for more updates on the rollout of this partnership to other markets.
In-market Audience segments now available in new markets
You can also reach the right audience through all of our In-market Audience segments, which we talked about back in November when seasonal segments were released in time for the holidays. The latest news with In-market Audiences is that they’re now available in open beta to all European markets (except for Belgium), so you can now reach people ready to buy in even more regions than before.
In addition to In-market Audiences, we’ve now also expanded the open beta of Customer Match to all markets (except for the European Union, UK and China), so you can target more high-value customers using your first-party data.
For more information on the open beta of these audience betas, contact Microsoft Advertising, and for a list of all the in-market audience segments available, check out our help page.
Lastly, in recent months we’ve improved our capabilities with Google Import so that you can now import In-market Audiences, remarketing and Custom Combination Lists from Google Ads (additional audience types like Customer Match will be added in the future). As part of this update improvement, we will now be updating all existing scheduled imports from Google Ads to enable them to import and update audience lists and associations. This enablement will happen by the end of January.
By the end of January, the Audience lists and targets option will be enabled to import In-market Audiences, remarketing and Custom Combination Lists from Google Ads.
New AI-based technologies to help boost your Microsoft Advertising performance
On the technology front, Microsoft Advertising has recently built a lot of powerful recommendations for your bids & budgets, keywords & targeting, and ads & extensions. In November we announced how you now have a one-stop shop for the recommendations in all your accounts, and there are now even more tactics to help you boost performance, such as examining what keywords are trending or where to set up conversion tracking.
Our AI-based recommendation engines have only gotten more effective with time, so we encourage you all to check out the Recommendations page to get your accounts off to the right start here in 2021.
Each recommendation includes the action you should take, how it’s generated, why it’s recommended, and the estimated performance gain to your manager account.
A sneak peek of what’s coming soon to Microsoft Advertising
As part of these monthly key product announcements, we now also want to provide you a brief preview of some other exciting technologies in Microsoft Advertising’s roadmap that you’ll be able to take advantage of soon. We’ll have more to share in future posts, but here’s a sneak peek of just a couple new capabilities:
* Coming soon to Microsoft Advertising’s automated bidding suite is a new strategy to help your brand get premium visibility and stay on top of the competition: Target Impression Share. This strategy will use artificial intelligence to automatically set bids at auction time with the goal of showing your ad on absolute top of the page, top of the page, or anywhere on the page.
Absolute top of the page: The very first ad that appears on the first search results page. Top of page: In the first set of ads that appear on the first search results page. Anywhere on the page: In any position on the first search results page.
* In the Recommendations page, you’ll soon be able to see your optimization score, which is a score that ranges from 0% to 100% to help indicate how much your account is performing at its full potential. When this is added, your recommendations and associated conversion impact will show how much the optimization score will be increased if you apply the recommendation, so you can better prioritize the actions you want to take.
Get your all-up optimization score and the estimated increase in optimization score across each of your recommendations tabs.
Optimization score is rolling out to some advertisers already, but more of you will be seeing this score at the account level and cross-account level in the future.
We hope these updates to help you reach the right audience with the right technology get you off to the right start in 2021, and we look forward to sharing more of these look-back and coming-soon updates throughout the rest of this year.
Take care and stay safe everybody!
Help us improve Microsoft Advertising
Your comments and feedback are integral to our ongoing efforts to shape and improve our product. For any questions or feedback, please reach out to your Microsoft Advertising account manager, or send us an email at advertising-feedback@microsoft.com. You can also interact with us on Twitter or suggest a feature on Microsoft Advertising Feature Suggestions.
https://j.mp/3s5O8Po
https://j.mp/3s5O8Po
Subscribe to:
Posts (Atom)


