A wire can be a bus if it is a serial link carrying many individual pieces of information. More usually, a bus is regarded as a collection of wires that transport digital information from A to B. 64 bit processors (PCs etc.) have a 64 bit-wide bus between the CPU and their memory chips and possibly to other devices.
It doesn't have to be inside a computer of course - anything that is transmitting information from A to B will use some form of wire or collection of wires for achieving those aims.
What differentiates a wire as not being a bus is that it only carries one coherent "entity" such as power or a microphone signal or is connected to an on/off switch or a guitar or a speaker. A bus is usually digital.
In ASK, a carrier is multiplied by a set of discrete amplitudes, depending on the information bits. In practice, binary ASP (BASK) is often used, where one of the amplitudes is zero, i.e. for a digital \$0\$, the modulated signal is zero, and for a digital \$1\$ the modulated signal is the carrier multiplied with some fixed amplitude. This important special case of ASK is called on-off keying (OOK).
Digital PAM is a digital modulation format where a pulse is multiplied by the current data symbol. In this sense, ASK can be seen as a special case of digital PAM, where the pulse is a sinusoid with the carrier frequency over one symbol interval. In general, the digital PAM signal can be written as
$$s(t)=\sum_{k=0}^{\infty}A_kp(t-kT)\tag{1}$$
where \$A_k\$ are the discrete symbols, \$p(t)\$ is the pulse function, and \$T\$ is the symbol interval. Note that \$A_k\$ can be any set of discrete symbols. Usually the number of symbols is a power of \$2\$. E.g., if the number of symbols is \$2^m\$, each symbol carries \$m\$ bits. Note that the signal \$s(t)\$ in (1) can be modulated by a carrier. In the general case, the symbols \$A_k\$ can be complex-valued, and a passband signal is generated by
$$\tilde{s}(t)=\Re\{s(t)e^{j\omega_c t}\}\tag{2}$$
where \$\omega_c\$ is the carrier frequency in radians. Equations (1) and (2) are the general representation of digital passband PAM. Quadrature amplitude modulation (QAM) and phase shift keying (PSK) are special cases.
Note that the type of PAM explained in nidhin's answer is analog PAM.
Best Answer
PAM encoding uses the physical amplitude of the sample as the final modulation (as seen on this page); i.e. it is an analogue modulation technique (the amplitude used for modulation is the actual sampled value, not the nearest approximation as is used in PCM, although it can be bounded).
Put another way, it is not (as part of the modulation) converted to a digital data word.
This modulation method forms the basis of QAM
When used as an analogue method, there is no inherent quantisation noise.
PCM takes the same thing a step further by requiring a step size (which may not be linear across the range) and then encoding that as a digital value (which will have a certain number of bits per sample); this will always introduce quantisation noise.
PCM data is widely used in the telephone system in most countries.