<p>Can anyone explain the huge difference in salary between UCLA’s CS and CSE programs?</p>
<p>CSE will have people venturing into the hardware side of things, which does not pay well. I guess making apps is more important than designing integrated circuits and microprocessors. Sigh.</p>
<p>
</p>
<p>I don’t think it’s a matter of importance, rather the market and revenue for applications is higher. Just like how NBA players make much more than surgeons.</p>
<p>Most companies need programmers. Not many companies need a computer engineer to build them a new microprocessor</p>
<p>National surveys usually show a higher average starting salary for hardware related majors than for CS. For example the Payscale report at <a href=“http://www.payscale.com/college-salary-report-2013/majors-that-pay-you-back[/url]”>http://www.payscale.com/college-salary-report-2013/majors-that-pay-you-back</a> lists average starting salaries as follows:</p>
<p>Electrical Engineering – $63,400
Computer Engineering – $62,700
Computer Science – $58,400</p>
<p>There are several reasons why UCLA might show very different data. It could simply relate to a small sample size. The survey only required a minimum of 5 participants per major to create an average and print results. This would explain some other odd results, such as Sociology majors having a higher starting salary than most engineering fields and more than double Psychology. However, Stanford shows a similar pattern, which makes me suspect that it more relates to location and having a lot of high salary CS-related jobs in the silicon valley area.</p>
<p>Re: computational linguistics, what the future holds I’m not sure. We have well over a dozen people doing voice recognition work and not one is a comp ling type. We buy the engine (the part that does most of the hard work) from one or two suppliers and use their tools to create our own ‘grammars’ i.e. vocabulary and so on. </p>
<p>It’s probably better if one looks for a job with a company that does engines and tools, rather than end users.</p>
<p>Thanks for the feedback! I’m still interested in UCLA’s “linguistics & computer science” degree but I’m starting to be disillusioned.</p>
<p>You better be. It is an extremely specialized field, to the tune that even those coworkers who left to work in other companies often end up doing voice related work. But it’s straight CS work, not linguistics per se. Our buddies at Nuance do the linguistics work.</p>
<p>With apologies to the rest of our readers, lemme give you some insight on how we end drones use voice recognition and text to speech (talk about thread hijacking :))</p>
<p>For text to speech it’s not THAT difficult. You basically have a string that goes “I would like some ice cream”. You initialize the TTS engine, make a few calls, and ask it to speak “I would like some ice cream”. There is a lot of linguistic work involved in ‘sculpting’ the speech so that it sounds natural, modern TTS’s can do miracles like “I am happy!!!” sounds a LOT different than “I am stuffed”. We’re given a set of control codes to embed the output so it sounds happy or upset or what not, i.e. “{happy}I am happy!!!” generates a happy tone, while “am I happy {question}” sounds like a question. Naturally this gets really ugly if you support multiple languages (a cubemate is doing Arabic TTS - quite funny". Other issues we have to deal with are words that are pronounced differently but spelled the same, pauses between utterances, etc.</p>
<p>Voice rec is much more complex. Typically we write a ‘grammar’ so that when we say “Play Artist Billy Joel” it comes back with a command code (Play Artist) and a parameter “Billy Joel”. This is easy, but we have a vocabulary of Siri size, and we can do it locally or at the cloud, so you always get back something recognized, and have to make sense of it. We do this “grammar” then the speech recognition engine recognizes the user said “such and such” and fires an event and tells us it recognized such and such. 99% of it is straightforward coding, no linguistic expertise. </p>
<p>Now that we’re moving more to a Siri model of free speech input (i.e. anything goes) there are techniques needed to handle such recognition. Again, we buy it from a vendor and/or use a cloud based recognizer. </p>
<p>The computational linguistics part is at the back end, the algorithms that can create grammars and recognize what was said with certain confidence. We simply call those as libraries or DLL’s or what nots. </p>
<p>The biggest and nextest thing will be semantic recognition, i.e. not just recognize what was said textually, but whether it makes sense semantically. Ironically, this was my graduate research in the mid 1980’s :)</p>
<p>If you can get a ‘real’ UCLA CS degree with a minor or double major in CL, that’s a different story.</p>
<p>
</p>
<p>Another possible reason is that there are many more weak (as in limited course offerings or less rigorous courses) computer science* degree programs out there than weak computer engineering or electrical engineering** degree programs out there. Perhaps the graduates of the weak computer science degree programs at schools where computer engineering or electrical engineering is absent are bringing down the overall computer science pay average, even though computer science graduates may out-earn computer engineering and electrical engineering graduates at schools that have all of them.</p>
<p>*Note that even generally prestigious schools may have weak computer science departments. Examples include Emory and Amherst.</p>
<p>**ABET accreditation imposes a higher floor on the quality of the degree programs for engineering majors. While ABET accreditation exists for computer science, it is not generally seen as necessary, so there are numerous non-ABET-accredited computer science degree programs, both good (e.g. Stanford) and bad (e.g. Emory and Amherst).</p>
<p>Why is UCLA’s CSE so little compared to UCLA’s CS… I’m going into CSE</p>
<p>Does anyone know the potential of UCSD’s “Mathematics & Comp Sci” major? And could I get a high paying job with that bachelor’s degree?</p>
<p>
</p>
<p>Doesn’t this dismiss the importance of accreditation? Those are three of the best CS schools in the world.</p>
<p>
</p>
<p>No it doesn’t dismiss it. Some fields just place less emphasis on accreditation than others. CS is one of those fields that just happens to place less emphasis on accreditation.</p>
<p>On the other hand, Aerospace Engineering and Mechanical Engineering are fields where accreditation is greatly preferred, if not required by employers/grad school.</p>
<p>… and seriously guys, quit squabbling over the differences between schools! If you had read the first three pages of this thread, you would understand that the major differences in pay are a result of cost of living. End of story.</p>
<p>Does anyone have any knowledge regarding my concern? (See above post)</p>