Location: Edinburgh, UK
Remote: Yes
Willing to relocate: No
Technologies: Machine learning, statistics, deep learning, data science, Python, PyTorch, Numpy, scikit-learn.
Résumé/CV: https://jmsrt.ch/docs/james_ritchie_cv.pdf
Email: hnhired@james.jmsrt.ch
Website: https://jmsrt.ch
I'm interested in machine learning and data science roles. I have expertise in machine learning and statistics having completed a PhD applying Bayesian inference to difficult scientific modelling problems using deep learning. I also have multiple years of experience working in startups as an engineer and data scientist across a range of industries.
Location: Edinburgh, UK
Remote: Yes
Willing to relocate: No
Technologies: Machine learning, statistics, deep learning, data science, Python, PyTorch, Numpy, scikit-learn.
Résumé/CV: https://jmsrt.ch/docs/james_ritchie_cv.pdf
Email: hnhired@james.jmsrt.ch
Website: https://jmsrt.ch
I'm interested in machine learning and data science roles. I have expertise in machine learning and statistics having completed a PhD applying Bayesian inference to difficult scientific modelling problems using deep learning. I also have multiple years of experience working in startups as an engineer and data scientist across a range of industries.
Location: Edinburgh, UK
Remote: Yes
Willing to relocate: No
Technologies: Machine learning, statistics, deep learning, data science, Python, PyTorch, Numpy, scikit-learn.
Résumé/CV: https://jmsrt.ch/docs/james_ritchie_cv.pdf
Email: hnhired@james.jmsrt.ch
Website: https://jmsrt.ch
I'm interested in machine learning and data science roles. I have expertise in machine learning and statistics having completed a PhD applying Bayesian inference to difficult scientific modelling problems using deep learning. I also have multiple years of experience working in startups as an engineer and data scientist across a range of industries.
Probabilistic Programming & Bayesian Methods for Hackers [1] by Cameron Davidson-Pilon is exactly what you want, starting from a computational-first perspective, then introducing the maths later, although it uses PyMC rather than Stan. It's freely available as a set of Jupyter notebooks, as well as a printed edition.
The LM computer on Apollo 14 had to be updated in flight in order to bypass a faulty abort switch.[1] The patch was applied by radioing the instructions to the crew and having them enter it manually.
Note that this was not a code patch: This was a sequence of changes to the program state that tricked it into ignoring the abort switch. The code itself was stored in read only core rope memory, so it was physically impossible to change it (the memory was written by appropriately winding wires around ferrite cores).
Yeah, but the The City is a tiny pebble in comparison to London. I'm sure there are plenty of places in London that size that are dominated by one industry (though none with their own laws.)
How is this different to SF (the 'city' in London isn't actually that big and smaller if you mean 'the square mile'), I assume most people commute to SF in as well?
It's not an apt comparison as firms are geographically distributed all over the bay area.
But people still commute there right? This is the point i'm trying to make, No one (and by that I man a small minority) lives in the city centre because it is too expensive. Applies to any major city on Earth which has suburbs.
I imagine the population during the day swells more than 100 fold...
The City of London is a weird ancient administrative division. It's not the whole of the city centre by any means, and not at all comparable to San Francisco.
I'm not sure you're aware of what's being discussed.
The Square Mile is an area of offices, mostly banks and company headquarters. It's the area of the ancient City of London, the bit the Romans walled off. There are now very few residential buildings. There aren't even any national government buildings, the area was controlled by merchants and guildsmen. Most people commute to work there, but you can say something similar about any 1mi² area in the most central bits of central London — few people live and work in Soho, for example.
Canada Square is in London's second financial district — Canary Wharf.
Most people working in London do live in London. The population of London (meaning the people who pay tax towards the public transport, vote for the mayor etc) is over 8 million.
Many startup companies are not far from the City of London / Square Mile — 5-10 minutes walk.
Canada Square is a long way away from the City of London (aka "Square Mile"). The City of London (that is, the small area to which the particular election protocol applies) has a tiny resident population; its resident:commuter ratio is not remotely comparable to that of London or SF.
Do IMSI catchers have any way of verifying the intercepted IMSIs are legitimate? If not, would it be possible to build a device to flood them with fake/spoofed IMSIs?
Technically, IMSI is not a device identifier. MT has IMEI and SIM (or more precisely ICC) has ICCID (which is normally never transmitted over the network). And the legislation probably specifically targets spoofing the IMEI, as spoofing IMSI does not gain you anything other than absence of service (on normal GSM/3GPP network).
From legal standpoint, device that spams IMSI catcher with registrations with random IMSIs is mostly same thing as the IMSI catcher itself, ie. device that requires it's own broadcast license to operate, as such device certainly does not meet legal (and technical) requirements for it to be an cellular phone.
On the other hand, generating random IMSI, burning that into ICC and thus producing unusable SIM is probably perfectly legal even when you put that inside normal GSM phone (from network standpoint it will behave mostly same as phone without any SIM). In practice SIMs with completely made-up IMSIs are even commercially available (idea there is that some phones will not fully boot without SIM).
No you won't, IMSI provisioning is mostly done today remotely. The SIM card will have an empty IMSI partition and once you've "activate" it with the parent network the network will provision an IMSI on that card.
If you have an account with a cell provider you'll carry the same IMSI number when you switch devices and SIM cards.
If you use pre-paid sim cards then those usually have thin provisioning of IMSI numbers they network buys a certain amount and activates them when the SIM card is activated and deactivates them once the SIM card has been inactive or not been topped off for a certain period of time (usually around 90 days).
When some one has your IMSI they can tie it directly to your personal details, phone number and various other details if they have sufficient access to the global cell system they can also get your location and what tower you are connected to.
Historical log data will give them any tower you've been connected too from the first tower during the initial activation and provisioning to the last tower you've been connected too.
Depending on the network and device most phones also send out nice diagnostic information about other towers and networks they see all the time regardless of what tower they are connected too at the time that with a given IMSI number will allow some one to triangulate the position of the device to under 10M in most urban areas if some one knows your IMSI they can pretty much pin point what room you are in at your house (give your house have several rooms pointing at different directions ;)).
And to be somewhat more pedantic, it even is not "unique device identifier" of the SIM, as SIM is not an device, but software application (typically one of applications) running on some hardware platform (usually an ICC, which is what is colloquially called "SIM card")
Like anything, it is technically possible to do this. As a defensive mechanism, I think it would make sense to perhaps hang out at city call and collects a bunch of IMSI's worth spoofing, and then rebroadcast those while you were being subject to surveillance, the goal being to force the adversary to follow down a bunch of leads which are bogus.
That said, this is completely illegal (unlike the actual surveillance which has been made legal by a series of unfortunate events). So I would not advise anyone to do this. Not to mention that unless you design your own cellular baseband radio circuit (so that you have access to all the docs) building something like this with "off the shelf" parts is quite expensive.
SDRs are no longer that expensive - even an Ettus or BladeRF device is within the range of the average middle-class engineer - and can be programmed completely in software to act as a cellular baseband device.
Yes, but the value of the collected data is much reduced if you can't distinguish between legitimate IMSIs and spoofed IMSIs of devices that were never actually there.
No, because he who collects isn't interested in random mobile phones, and whoever tries to spoof "other" phones doesn't know the phones of interest of the collector. And the collector is interested not in knowing the presence but in the whole traffic, so the spoofing is even more obvious.
If you attach to a network successfully, it means the its secret key stored in the SIM card matches that of the network.
Since your IMSI catcher is basically just a proxy between the mobile phone and the real network, you can therefore easily detect that the IMSI is who it says it is.
Junior engineer with an interest in data science, additional experience in backend development. Ideally looking for teams with interesting problems they want to solve with machine learning.
