Research Diary

For the longest time I am waiting for someone to create an USB key format network adapter that has builtin firewall, antivirus and VPN software and is activated by fingerprint authentication. For corporations this kind of device would help a lot with edge control and also help preventing outbreaks when non-corporate devices are attached to the network via office networks or VPN. These devices should have corresponding central server on the corporate site to connect and download policies, rules and updates and upload reports regularly. Unlike software based tools this could work with various OSes and virtual OSes running within the OS cannot circumvent the network monitoring. Dedicated chipsets can increase performance and user "errors" becomes a non-issue. If the user does not engage the USB key for network access then corporate access is not allowed and system logs this to report at the next corporate connection. It is kind of big brother type of control but lately the threat levels are getting higher and higher for both the corporations and their clients.

Check out the new pictures of Lara. Picture were taken a day before she turned 2.

Finally found some time and upgraded to Wordpress 2.5. No major changes on the front-end but the back-end has been definitely improved. I am planning to add some more plugins for PicasaWeb and RSS feeds in the next couple of weeks.

Okay everybody is talking about Biofuels. All big companies are investing in biofuels. Some do it to look nice and use it as a PR stunt (most oil companies), knowing that the current attempts will not even make a dent in their market share or profits, some others do it because it might be profitable in the short term locally, but globally I need to see proof of its feasibility. First of all you need to build either hundreds of large (very very large) plants to process the biowastes, or crops. You need to transfer them in (constantly), you need to store them, process them, ferment them in huuuggeee tanks, then filter them, purify them, get rid of the waste and transfer the product to the market. You will need to control these plants very gently, since you will be using organisms you need to make sure temperatures do not swing much (it gets cold in midwest where the crops are easy to get), you have to maintain and mix carefully to keep things going (either huge tanks or huge number of tanks…). Or you can build thousands of smaller plants that can handle local crops. Then capital costs will skyrocket. Even if you can streamline the plants by standardizing on parts, can you imaging recalling 200,000 biofuel plants because of faulty valves? Also grass grows for half of the year, what are they going to use for the rest of the year? At some point you will need to tap into the food stock. It will increase the food costs, which in affect will have a worse effect on low income population then high fuel costs. You can do without driving but you cannot do without eating. As I said somebody has to show me in clear math how this would work.

I would be more interested using garbage and landfills as organic source for biofuels. It is produced constantly. There is an abundance of it. Nobody needs it, everybody is trying to get rid of it. You can use that and the CO2 from the air to convert it to biofuels… Less garbage, less CO2, less dangerous wastes and you also get biofuels. I even have the name for the company: GARBAGAS!

Update: Apparently there is already a company doing this. Not using biotech but old style chemical engineering techniques:

Update 2: Okay there is more news on this topic. More people are looking into alternative sources for biofuels. Instead of using/abusing corn more research is done on creating gasoline directly from other types of wastes. Check new here and here.

I think in the next 18 months there will be a very big change of direction in Biofuel research. Especially change in administration might change policy. I also expect big oil companies getting little more nervous about the new high throughput alternatives.

Tidbits from behavioral economics: What counts as a gain or a loss is relative to prior prices, not considered as independent of an investor’s attention. A stock acquired at $50 and and acheives a year-end value of $100. A few months into the new year it is sold at $75. While this is objectively a $25 gain, the investor likely feels as if there was a loss. One classic decision-making bias Nobel Laureate Daniel Kahneman and Amos Tversky identifed is this effect, called anchoring and adjustment.

Operon: A unit of genetic material that functions in a coordinated manner by means of an operator, a promoter, and one or more structural genes that are transcribed together.

Sigma Factor: A sigma factor (σ factor) is a prokaryotic transcription initiation factor that must be part of RNA polymerase (RNAP) for specific binding to promoter sites on DNA. Different sigma factors are activated in response to different environmental conditions, as are listed below. Every molecule of RNA polymerase contains exactly one sigma factor subunit, which in the model bacterium Escherichia coli is one of those listed below. E.coli has at least eight sigma factors; the number of sigma factors varies between bacterial species. All sigma factors are distinguished by their characteristic molecular weights. For example, σ70 refers to the sigma factor with a molecular weight of 70 kDa.

It has been a while since I could post something here. Life has been busy and I was working on some ideas. At MIT I am working on some scoring algorithms to rank Antimicrobial Peptides to design strong ones later. For the timeseries prediction project I was developing some framework ideas that I will start testing soon. At work I am working on the new Sharepoint 2007 Server and some more interesting virtualization projects. Lara is growing nicely. I am attaching some more new pictures of here below.

After a move and machine setup I am back again. I am planning to post more often than before after most probably mid September. I am working on two projects both related to pattern recognition. More to follow soon. I am also moving most of my pictures to Picasa Web. Once the move is completed I will start posting more…

This post and its comments will include information that I will collect on Hidden Markov Chains.